CD3 ANTIBODY AND PHARMACEUTICAL USE THEREOF

Abstract

 The present disclosure relates to a CD3 antibody and a pharmaceutical use thereof. Specifically, the present disclosure relates to forming a multi-specificity antibody by using CD3 antibody and binding molecules of another target. The multi-specificity antibody may simultaneously bind to CD3 and another tumor-associated antigen, and bind and activate CD3-positive T cells while binding tumor-associated antigen-expressing cells, thereby promoting T cells specifically killing tumor cells that express tumor-associated antigens. In addition, the present disclosure also provides a preparation and application of a multi-specific antibody.

Description

FIELD OF THE INVENTION

[0001] The present invention relates to CD3 antibodies such as humanized CD3 antibodies, and multispecific antibodies simultaneously binding to CD3 and tumor-associated antigen(s).

BACKGROUND OF THE INVENTION

[0002] The descriptions herein only provide background information about the present invention, and do not necessarily constitute prior art.

[0003] CD3 is a T cell co-receptor composed of four different chains (Wucherpfennig, KW et al. (2010) Structural Biology of The T cell Receptor: Insights Into Receptor Assembly, Ligand Recognition, and Initiation of Signaling, Cold Spring Harb. Perspect. Biol. 2(4):A005140; pages 1-14; Chetty, R. et al. (1994) CD3:Structure, Function, And Role Of Immunostaining In Clinical Practice, J. Pathol. 173(4):303-307; Guy, C.S. et al. (2009) Organization of Proximal Signal Initiation at the TCR:CD3 Complex, Immunol.Rev.232(1):7-21).

[0004] In mammals, the complex formed by CD3 multi-subunits is associated with T cell receptor (TCR) molecule to generate activation signals in T lymphocytes (Smith-Garvin, JE et al. (2009) T Cell Activation, Annu. Rev. Immunol. 27:591-619). In the absence of CD3, TCR neither can be assembled properly nor be degraded (Thomas, S. et al. (2010) Molecular Immunology Lessons From Therapeutic T cell Receptor Gene Transfer, Immunology 129(2): 170-177). Studies have found that CD3 binds to the membranes of all mature T cells and hardly binds to other cell types (Janeway, C.A. et al. (2005): Immunobiology: The Immune System in Health and Disease, 6th Edition, Garland Science Publishing, NY, pp.214-216; Sun, Z.J. et al. (2001) Mechanisms Contributing to T Cell Receptor Signaling and Assembly Revealed by the Solution Structure of an Ectodomain Fragment of the CD3ε: γ Heterodimer, Cell 105 (7): 913-923; Kuhns, M.S. et al. (2006) Deconstructing the Form and Function of the TCR/CD3 Complex, Immunity. 2006 Feb, 24(2): 133-139).

[0005] The constant CD3ε signaling component of the T cell receptor (TCR) complex on T cells has been used as a target to promote the formation of immunological synapses between T cells and tumor cells. The co-engagement of CD3 and tumor antigen(s) activates T cells, resulting in the lysis of tumor cells expressing tumor antigen(s) (Baeuerle et al. (2011) Bispecific T Cell Engager for Cancer Therapy, In: Bispecific Antibodies, Kontermann, R.E.(Ed.) Springer-Verlag; 2011:273-287). This method allows bispecific antibodies to fully interact with T cell compartments with a specificity which is as high as that for tumor cells, and the method is widely applicable to a variety of cell surface tumor antigens.

[0006] B7H3 is a member of the B7 family and belongs to the type I transmembrane protein. It contains a signal peptide at the amino terminus, an extracellular immunoglobulin-like variable region (IgV), a constant region (IgC), a transmembrane region and a cytoplasmic tail region comprising 45 amino acids (Tissue Antigens. 2007 Aug; 70 (2): 96-104). Currently, B7H3 mainly has two types of splicing forms, B7H3a and B7H3b. The extracellular segment of B7H3a is composed of two immunoglobulin domains, IgV-IgC, also known as 2IgB7H3, whereas the extracellular segment of B7H3b is composed of four immunoglobulin domains, IgV-IgC-IgV-IgC, also known as 4IgB7H3.

[0007] B7H3 protein is absent in normal tissues and cells, or expressed in normal tissues and cells at extremely low level; however, it is highly expressed in a variety of tumor tissues, and is closely related to tumor progression, patient survival and prognosis. It has been clinically reported that B7H3 is over-expressed in various cancer types, especially in non-small cell lung cancer, kidney cancer, urinary tract epithelial cancer, colorectal cancer, prostate cancer, glioblastoma multiforme, ovarian cancer and pancreatic cancer. (Lung Cancer. 2009 Nov; 66(2): 245-249; Clin Cancer Res. 2008 Aug 15; 14(16): 5150-5157). In addition, it has also been reported in literatures that the expression intensity of B7H3 is positively correlated with clinically pathological malignancy in prostate cancer (such as tumor volume, invasion beyond the prostate or Gleason score), and is also correlated with cancer progression (Cancer Res. 2007 Aug 15; 67 (16): 7893-7900). Similarly, the expression of B7H3 is negatively correlated with event-free survival in glioblastoma multiforme, and the expression of B7H3 is correlated with lymphatic metastasis and pathological progression in pancreatic cancer. Therefore, B7H3 is considered as a new tumor marker and potential therapeutic target.

SUMMARY OF THE INVENTION

[0008] The present disclosure provides antibodies or antigen-binding fragments thereof that can specifically bind to human CD3.

[0009] In one aspect, the present disclosure provides an antibody or antigen-binding fragment thereof specifically binding to human CD3, the antibody or antigen-binding fragment thereof comprising a heavy chain variable region and a light chain variable region, wherein the light chain variable region comprises LCDR1, LCDR2, and LCDR3 as shown in SEQ ID NOs: 48, 49, and 50, respectively, and
the heavy chain variable region is any one selected from the group consisting of the following i) to v):

i) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 38 and 39, respectively;

ii) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 40 and 41, respectively;

iii) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 40 and 42, respectively;

iv) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 40 and 43, respectively; and

v) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 47 and 45, respectively.

[0010] In some embodiments, the antibody or antigen-binding fragment thereof specifically binding to human CD3 is a humanized antibody or antigen-binding fragment thereof.

[0011] In some embodiments, the antibody or antigen-binding fragment thereof specifically binding to human CD3 comprises a light chain variable region of SEQ ID NO: 36 and/or a heavy chain variable region as shown in any one selected from the group consisting of SEQ ID NOs: 29, 30, 31, 32 and 35.

[0012] In some embodiments, the antibody or antigen-binding fragment thereof specifically binding to human CD3 further comprises an antibody light chain constant region and/or a heavy chain constant region; optionally, the light chain constant region is a light chain constant region of a human kappa, lambda chain or variant thereof, and the heavy chain constant region is a heavy chain constant region of a human IgG1, IgG2, IgG3, IgG4 or variant thereof.

[0013] In some embodiments, the antigen-binding fragment is selected from Fab, Fab', F(ab')2, dimerized V region (diabody) and disulfide-stabilized V region (dsFv).

[0014] In another aspect, the present disclosure provides a single-chain antibody comprising the light chain variable region and the heavy chain variable region of the antibody or antigen-binding fragment thereof specifically binding to human CD3 as described above.

[0015] In some embodiments, the sequence of the single-chain antibody is as shown in SEQ ID NO: 55, 56, 57, 58, 61, 62, 63, 64, 65 or 68.

[0016] In another aspect, the present disclosure provides a multispecific antibody specifically binding to human CD3 and tumor-associated antigen(s) (TAA), the multispecific antibody comprising the single-chain antibody or the antibody or antigen-binding fragment thereof specifically binding to human CD3 as described above.

[0017] In some embodiments of the multispecific antibody, the tumor-associated antigen is selected from the group consisting of AFP, ALK, B7H3, BAGE protein, BCMA, BIRC5(survivin), BIRC7, β-catenin, brc-ab1, BRCA1, BORIS, CA9, CA125, carbonic anhydrase IX, caspase -8, CALR, CCR5, CD19, CD20(MS4A1), CD22, CD30, CD33, CD38, CD40, CD123, CD133, CD138, CDK4, CEA, Claudin 18.2, cyclin-Bl, CYP1B1, EGFR, EGFRvIII, ErbB2/Her2, ErbB3, ErbB4, ETV6-AML, EpCAM, EphA2, Fra-1, FOLR1, GAGE protein (such as GAGE-1, -2), GD2, GD3, GloboH, Glypican-3, GM3, gp100, Her2, HLA/B-raf, HLA/k-ras, HLA/MAGE-A3, hTERT, IL13Rα2, LMP2, κ-Light, LeY, MAGE protein (such as MAGE-1, -2, -3, -4, -6 and -12), MART-1, mesothelin, ML-IAP, MOv-y, Muc1, Muc2, Muc3, Muc4, Muc5, Muc16(CA-125, MUM1, NA17, NKG2D, NY-BR1, NY-BR62, NY-BR85, NY-ESO1, OX40, p15, p53, PAP, PAX3, PAX5, PCTA-1, PLAC1, PRLR, PRAME, PSMA (FOLH1), RAGE protein, Ras, RGS5, Rho, ROR1, SART-1, SART-3, STEAP1, STEAP2, TAG-72, TGF-β, TMPRSS2, Thompson-nouvelle antigen (Tn), TRP-1, TRP-2, tyrosinase, uroplakin-3 and 5T4 (Trophoblast glycoprotein). Preferably, the tumor-associated antigen is selected from the group consisting of B7H3, BCMA, CEA, CD19, CD20, CD38, CD138, Claudin 18.2, PSMA and mesothelin.

[0018] In another aspect, the present disclosure provides a pharmaceutical composition comprising a therapeutically effective amount of the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single chain antibody, or the multispecific antibody as described above, and one or more pharmaceutically acceptable carriers, diluents, buffers or excipients. In some embodiments, the therapeutically effective amount means that the composition comprises a unit dose of 0.1-3000 mg (more preferably 1-1000 mg) of the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, or the multispecific antibody as described above.

[0019] In another aspect, the present disclosure provides an isolated nucleic acid molecule, encoding the antibody or antigen-binding fragment thereof specifically binding to human CD3, encoding the single-chain antibody, or encoding the multispecific antibody as described above.

[0020] In another aspect, the present disclosure provides a recombinant vector comprising the isolated nucleic acid molecule as described above.

[0021] In another aspect, the present disclosure provides a host cell transformed with the aforementioned recombinant vector, and the host cell is selected from prokaryotic cell and eukaryotic cell, preferably eukaryotic cell, more preferably mammalian cell or insect cell.

[0022] In another aspect, the present disclosure provides a method for producing the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, or the multispecific antibody as described above, the method comprising culturing the host cell as described above in a culture medium to form and accumulate the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, or the multispecific antibody as described above, and recovering the antibody or antigen-binding fragment thereof, the single-chain antibody or the multispecific antibody from the culture.

[0023] In another aspect, the present disclosure provides the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, the multispecific antibody, the pharmaceutical composition, or the isolated nucleic acid molecule as described above, as a medicament. In some embodiments, the medicament is a medicament for the activation of T cells; In some embodiments, the medicament is a medicament for the treatment of cancer, autoimmune disease or inflammatory disease.

[0024] In another aspect, the present disclosure provides use of the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, the multispecific antibody, the pharmaceutical composition, or the isolated nucleic acid molecule as described above in the preparation of a medicament for the activation of T cells.

[0025] In another aspect, the present disclosure provides a method for activating T cells, the method comprising administering to a subject a therapeutically effective amount of the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, the multispecific antibody, the pharmaceutical composition, or the isolated nucleic acid molecule as described above. In some embodiments, the method comprises administering to a subject a composition comprising a unit dose of 0.1-3000 mg (more preferably 1-1000 mg) of the multispecific antibody as described above, or the pharmaceutical composition, or the isolated nucleic acid molecule as described above.

[0026] In another aspect, the present disclosure provides a method for the treatment of cancer, autoimmune disease or inflammatory disease, the method comprising administering to a subject a therapeutically effective amount of the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, the multispecific antibody, the pharmaceutical composition, or the isolated nucleic acid molecule as described above. In some embodiments, the method comprises administering to a subject a composition comprising a unit dose of 0.1-3000 mg (more preferably 1-1000 mg) of the multispecific antibody as described above, or the pharmaceutical composition, or the isolated nucleic acid molecule as described above.

[0027] In another aspect, the present disclosure provides use of the antibody or antigen-binding fragment thereof specifically binding to human CD3, the single-chain antibody, the multispecific antibody, the pharmaceutical composition, or the isolated nucleic acid molecule as described above in the preparation of a medicament for the treatment of cancer, autoimmune disease or inflammatory disease.

[0028] In some embodiments, the cancer described above is any one selected from the group consisting of carcinoma, lymphoma, blastoma, sarcoma, leukemia and lymphoid malignancies. More specific examples of the cancer include squamous cell carcinoma, myeloma, small cell lung cancer, non-small cell lung cancer (NSCLC), head and neck squamous cell carcinoma (HNSCC), glioma, Hodgkin's lymphoma, Non-Hodgkin's lymphoma, diffuse large B-cell lymphoma (DLBCL), follicular lymphoma, acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), primary mediastinal large B-cell lymphoma, mantle cell lymphoma (MCL), small lymphocytic lymphoma (SLL), T-cell/histocyte-rich large B-cell lymphoma, multiple myeloma, myeloid leukemia-protein 1 (Mcl-1), myelodysplastic syndrome (MDS), gastrointestinal (tract) cancer, kidney cancer, ovarian cancer, liver cancer, lymphoblastic leukemia, lymphocytic leukemia, colorectal cancer, endometrial cancer, kidney cancer, prostate cancer, thyroid cancer, melanoma, chondrosarcoma, neuroblastoma, pancreatic cancer, glioblastoma multiforme, gastric cancer, bone cancer, Ewing sarcoma, cervical cancer, brain cancer, gastric cancer, bladder cancer, hepatocellular tumor, breast cancer, colon cancer, hepatocellular cancer (HCC), clear cell renal cell carcinoma (RCC), head and neck cancer, pharyngolaryngeal cancer, hepatobiliary cancer, central nervous system cancer, esophagus cancer, malignant pleural mesothelioma, systemic light chain amyloidosis, lymphoplasmacytic lymphoma, myelodysplastic syndrome, myeloproliferative tumor, neuroendocrine tumor, Merkel cell cancel, testicular cancer, and skin cancer. In some embodiments, the cancer is B7-H3 positive cell-related cancer; preferably breast cancer, ovarian cancer, prostate cancer, pancreatic cancer, kidney cancer, lung cancer, liver cancer, gastric cancer, colon cancer, bladder cancer, esophagus cancer, cervical cancer, gallbladder cancer, glioblastoma or melanoma.

[0029] In some embodiments, the autoimmune disease or inflammatory disease described above is any one selected from the group consisting of rheumatoid arthritis, psoriasis, Crohn's disease, ankylosing spondylitis, multiple sclerosis, type I diabetes, hepatitis, myocarditis, Sjogren syndrome, autoimmune hemolytic anemia after transplant rejection, vesicular pemphigoid, Graves disease, Hashimoto thyroiditis, systemic lupus erythematosus (SLE), myasthenia gravis, pemphigus and pernicious anemia.

DESCRIPTION OF THE DRAWINGS

[0030] 

Figure 1A and Figure 1B: Figure 1A is a schematic diagram of a bivalent bispecific antibody, and Figure 1B is a schematic diagram of a monovalent bispecific antibody.

Figure 2A to Figure 2D: Detection of the antibodies by flowcytometry for the activity of binding to cells with the expression of or without the expression of corresponding antigen. Figure 2A shows the detection of the activity of various antibodies to bind to A498 cells expressing human B7H3; Figure 2B shows the detection of the activity of various antibodies to bind to CT26 cells over-expressing human B7H3; Figure 2C shows the detection of the activity of various antibodies to bind to CT26 cells which do not express human B7H3, and the results show that none of the antibodies bind to CT26 cells that do not express human B7H3; Figure 2D shows the detection of the activity of various antibodies to bind to Jurkat recombinant cells expressing CD3. The vertical axis in Figures 2A to Figure 2D represents the geometric mean of the fluorescence signal.

Figure 3A to Figure 3B: Detection of the activity of bispecific antibodies comprising various CD3 scFvs in killing A498. Figure 3A shows the killing activity of the B7H3 monovalent bispecific antibody. Figure 3B shows the killing activity of the B7H3 bivalent bispecific antibody. All of the bispecific antibodies show obvious killing activity, regardless of B7H3 monovalent or bivalent, except 155, 156, 185 and 186 which have weaker killing activity against A498.

Figure 4A to Figure 4B: Comparison of the killing activity against A498 between B7H3 mono- and bi-valent bispecific antibodies comprising the same CD3 scFv. Figure 4A shows a comparison of the killing activity between the B7H3 monovalent (181) and bivalent (131) bispecific antibodies comprising HRH1. Figure 4B shows a comparison of the killing activity between the B7H3 monovalent (187) and bivalent (177) comprising HRH7. All of the experimental results show that the B7H3 bivalent bispecific antibodies have more obvious A498 killing activity compared to the B7H3 monovalent bispecific antibodies. At the same time, the B7H3 bivalent bispecific antibodies have significantly stronger killing activity than that of B7H3 monovalent bispecific antibodies.

Figures 5A to Figure 5C: Detection of the killing activity against A498 of B7H3 bivalent bispecific antibodies comprising the same CD3 heavy chain variable region, but with different arrangement orders. Figure 5A shows a comparison of the killing activity of the B7H3 bivalent bispecific antibodies comprising a first polypeptide chain comprising HRH2, wherein the first polypeptide chain is arranged in various orders (AFF1, AFF2, AFF3, AFF4). Figure 5B shows a comparison of the killing activity of the B7H3 bivalent bispecific antibodies comprising a second polypeptide chain comprising HRH2, wherein the second polypeptide chain is arranged in various orders (AFF3, AFF3-B). The results show that all the B7H3 bivalent bispecific antibodies with the same sequence but with different arrangement of VH and VL have significant A498 cell killing activity, and the molecules with different arrangement orders have similar killing activity. Figure 5C shows a comparison of the killing activity between the bispecific antibodies comprising the same B7H3 scFv and CD3 scFv, but with different structures. The three test bispecific antibodies 127, 201 and 202 all have the ability to kill A498 tumor cells, of which the bispecific antibody 127 exhibits killing activity superior to that of 201 and 202.

Figure 6A to Figure 6B: Detection of the activation of Jurkat recombinant cells by different antibodies. Figure 6A shows the antibody-mediated B7H3 target-specific activation of Jurkat recombination cell, in the presence of A498 cells; Figure 6B shows the antibody-mediated non-B7H3 target-specific activation of Jurkat recombination cell, in the absence of A498 cells. The same antibody legends are indicated in Figure 6A to Figure 6B.

Figure 7A to Figure 7B: Detection of the activation of Jurkat recombinant cells by the bispecific antibodies comprising the same CD3scFv, but with different valences. Figure 7A shows the antibody-mediated B7H3 target-specific activation of Jurkat recombination cell by the B7H3 mono/bivalent bispecific antibodies in the presence of A498 cells;

Figure 7B shows the antibody-mediated non-B7H3 target-specific activation of Jurkat recombination cell by the B7H3 mono/bivalent bispecific antibodies in the absence of A498 cells.

Figure 8A to Figure 8C: Various antibodies are tested, in the presence of A498 cells, for the stimulation of PBMCs to produce B7H3 target-specific cytokine secretion. Figure 8A shows a comparison of IFNγ secretion levels from PBMCs stimulated by various antibodies; Figure 8B shows a comparison of TNFα secretion levels from PBMCs stimulated by various antibodies; and Figure 8C shows a comparison of IL-2 secretion levels from PBMCs stimulated by various antibodies. Figure 8A- Figure 8C show that antibodies 118, 127 and 132 can significantly stimulate PBMC to produce B7H3 target-specific cytokine secretion. The same antibody legends are indicated in Figure 8A to Figure 8C.

Figure 9A to Figure 9C: Various antibodies are tested, in the presence of CHOK1 cells (without the expression of B7H3), for the stimulation of PBMCs to produce non B7H3 target-specific cytokine secretion. Figure 9A shows a comparison of IFNγ levels secreted from PBMCs stimulated by various antibodies; Figure 9B shows a comparison of TNFα levels secreted from PBMCs stimulated by various antibodies; and Figure 9C shows a comparison of IL-2 levels secreted from PBMCs stimulated by various antibodies. Figure 9A- Figure 9C show that antibodies 118, 127, and 132 cannot stimulate PBMC to produce non B7H3 target-specific cytokine secretion, and have better safety. The same antibody legends are indicated in Figure 9A to Figure 9C.

Figure 10A to Figure 10E: Detection of the anti-tumor efficacy of the bispecific antibodies in mouse A498 model reconstituted with human PBMC. Figure 10A shows detection of the anti-tumor activity of the low-dose B7H3 bivalent bispecific antibodies. Both the low-dose antibodies 118 and 119 still show certain anti-tumor activity and show a certain degree of dose-dependency. Figure 10B shows detection of the anti-tumor activity of the B7H3 bivalent bispecific antibodies at doses of 0.3 mpk and 0.6 mpk. The antibody 113 shows an in vivo dose-dependent tumor inhibitory activity. Figure 10C shows detection of the anti-tumor activity of the B7H3 bivalent bispecific antibodies at doses of 0.12mpk and 0.36mpk. Antibody 118 shows significant anti-tumor activity at both doses. Figure 10D shows detection of the anti-tumor activity of the B7H3 bivalent bispecific antibodies at a dose of 0.36 mpk. Antibodies 126, 127 and 128 all show significant anti-tumor activity. Figure 10E shows the anti-tumor activity of antibody 127 at various doses and various dosing frequencies. In Figure 10A to Figure 10E, "Vehicle" represents a negative control group administered with PBS.

Figure 11A to Figure 11B: Anti-tumor efficacy of the bispecific antibodies in hCD3 KI mouse model. Figure 11A and Figure 11B show the anti-tumor effects of 118 and 132 in hCD3 KI mouse model, respectively.

DETAILED DESCRIPTION OF THE INVENTION

Terminology (Definition)

[0031] Three-letter codes and one-letter codes for amino acids used in the present disclosure are as described in J. biol. chem, 243, p3558(1968).

[0032] The term "multispecific protein molecule" refers to a protein molecule capable of specifically binding to two or more target antigens or target antigen epitopes. A protein molecule that can specifically bind to two target antigens or target antigen epitopes is named as bispecific protein molecule, including antibodies or antigen-binding fragments of the antibodies (such as single-chain antibodies). "Bispecific protein molecule" is interchangeable with "bispecific antibody" herein.

[0033] The term "binding region" for an antigen refers to a region or part that can specifically bind to an antigen in a multispecific protein molecule or antibody molecule. The antigen-binding region can be a ligand binding domain that can directly bind to the antigen, or can be a domain comprising a variable region of the antibody that can directly bind to the antigen.

[0034] The term "antibody (Ab)" includes any antigen binding molecule or molecular complex that includes at least one complementarity determining region (CDR) that specifically binds to or interacts with a specific antigen (e.g., CD3). The term "antibody" includes: four polypeptide chains connected to each other via disulfide bond(s), immunoglobulin molecules comprising two heavy (H) chains and two light (L) chains, and multimers thereof (e.g., IgM). Each heavy chain includes a heavy chain variable region (hereinafter abbreviated as HCVR or VH) and a heavy chain constant region. This heavy chain constant region comprises three regions (domains): CHI, CH2 and CH3. Each light chain includes a light chain variable region (hereinafter abbreviated as LCVR or VL) and a light chain constant region. The light chain constant region comprises one region (domain, CL1). VH and VL regions can be further subdivided into hypervariable regions, named as complementarity determining regions (CDRs), among which the more conservative regions are interspersed, named as framework regions (FRs). Each VH and VL are composed of three CDRs and four FRs, arranged from the amino terminus to the carboxyl terminus in the following order: FR1, CDR1, FR2, CDR2, FR3, CDR3, FR4. In various embodiments of the present disclosure, the FRs of the anti-CD3 antibodies (or antigen-binding portions thereof), anti-B7H3 antibodies (or antigen-binding portions thereof), or antibodies against other target antigens can be the same as the human germline sequence, or can be naturally or artificially modified. The antibodies can be antibodies of different subclasses, for example, IgG (e.g., IgG1, IgG2, IgG3 or IgG4 subclass), IgA1, IgA2, IgD, IgE, or IgM antibody.

[0035] The term "antibody" also encompasses antigen-binding fragments of the full antibody molecules. The terms "antigen-binding portion", "antigen-binding domain", "antigen-binding fragment", etc. of an antibody, as used herein, include any naturally occurring, enzymatically produced, synthetic or genetically engineered polypeptide or glycoprotein that specifically binds to an antigen to form a complex. Antigen-binding fragments of an antibody can be derived from, for example, the full antibody molecule by using any suitable standard technique, such as proteolytic digestion or recombinant genetic engineering techniques involving manipulation and expression of DNA encoding antibody variable regions and (optionally) constant regions. The DNA is known and/or can be easily obtained from, for example, commercially available sources, DNA databases (including, for example, phage-antibody databases), or can be synthesized. The DNA can be sequenced and manipulated chemically or by using molecular biotechnology, for example, by arranging one or more variable and/or constant regions into a suitable configuration, or by introducing codons, generating cysteine residues, modification, additon or deletion of amino acids, etc.

[0036] Non-limiting examples of antigen-binding fragments include: (i) Fab fragment; (ii) F(ab')2 fragment; (iii) Fd fragment; (iv) Fv fragment; (v) single-chain Fv (scFv) molecule; (vi) dAb fragment. Other engineered molecules, such as region-specific antibodies, single-domain antibodies, region-deleted antibodies, chimeric antibodies, CDR-grafted antibodies, diabodies, tribodies, tetrabodies, minibodies, nanobodies (e.g. monovalent nanobodies, bivalent nanobodies, etc.), Small Modular Immunopharmaceuticals (SMIP) and Shark Variable IgNAR regions are also included in the term "antigen-binding fragment" as used herein.

[0037] The antigen-binding fragment of an antibody will typically contain at least one variable region. The variable region can be a region of any size or amino acid composition and will generally comprise one or more CDRs adjacent to or within the framework sequences. In an antigen-binding fragment having a VH region associated with a VL region, the VH and VL regions can be located opposite to each other in any suitable arrangement. For example, the variable region can be dimerized and comprises VH-VL or VL-VH dimers.

[0038] In some embodiments, the antigen-binding fragment of the antibody is in any configuration of variable region and constant region, the variable region and the constant region can be directly connected to each other or can be connected through a complete or partial hinge or linker region. The hinge region can be composed of at least 2 (for example, 5, 10, 15, 20, 40, 60 or more) amino acids, so that a flexible or semi-flexible connection will be generated between the adjacent variable and/or constant regions in a single polypeptide molecule. Furthermore, the antigen-binding fragments of the antibodies of the present invention can include homodimers or heterodimers (or other multimers) in any configuration of variable region and constant region as indicated above, wherein the variable region and constant region can be non-covalently connected to each other and/or connected to one or more monomeric VH or VL regions (e.g., via disulfide bond(s)).

[0039] "Murine antibody" as used herein refers to mouse-derived monoclonal antibodies prepared according to the knowledge and skills in the art. During the preparation, test subjects are injected with an antigen, and then a hybridoma expressing the antibody which possesses desired sequence or functional characteristics is isolated. The resulting antibody will be a murine antibody if the injected test subjects are mice.

[0040] The "chimeric antibody", is an antibody by fusing the variable region of murine antibody together with the constant region of human antibody, and such antibody can alleviate the murine antibody-induced immune response. To establish a chimeric antibody, a hybridoma secreting specific murine monoclonal antibody is established firstly, and variable region gene is cloned from the murine hybridoma. Then constant region gene is cloned from human antibody according to the need. The murine variable region gene is connected to the human constant region gene to form a chimeric gene, which can be subsequently inserted into an expression vector. Finally the chimeric antibody molecule will be expressed in eukaryotic or prokaryotic system. In a preferable embodiment of the present disclosure, the antibody light chain of the chimeric antibody further comprises a light chain constant region of a human kappa, lambda chain or variant thereof. The antibody heavy chain of the chimeric antibody further comprises a heavy chain constant region of human IgG1, IgG2, IgG3, IgG4 or variant thereof, preferably comprises a heavy chain constant region of human IgG1, IgG2 or IgG4, or comprises a heavy chain constant region variant of human IgG1, IgG2 or IgG4 with amino acid mutations (such as YTE mutation or back mutation, L234A and/or L235A mutation, or S228P mutation).

[0041] The term "humanized antibody", including CDR-grafted antibody, refers to an antibody generated by grafting animal-derived antibody, e.g., murine antibody CDR sequences into human antibody variable region frameworks (i.e., framework regions). Humanized antibodies can conquer heterologous responses induced by chimeric antibodies which carry a large number of heterologous protein components. Such framework sequences can be obtained from public DNA database covering germline antibody gene sequences or published references. For example, germline DNA sequences of human heavy and light chain variable region genes can be found in "VBase" human germline sequence database (available on http://www.vbase2.org/), as well as in Kabat, EA, et al. 1991 Sequences of Proteins of Immunological Interest, 5th Ed. To avoid a decrease in activity caused by the decreased immunogenicity, the framework sequences in human antibody variable region can be subjected to minimal reverse mutations or back mutations to maintain the activity. The humanized antibody of the present disclosure also comprises humanized antibody on which CDR affinity maturation is performed by phage display.

[0042] Due to the residues contacted with an antigen, the grafting of CDR can result in a decreased affinity of an antibody or antigen binding fragment thereof to the antigen due to the framework residues contacted with the antigen. Such interactions can be resulted from highly somatic mutations. Therefore, it may still be necessary to graft the donor framework amino acids onto the humanized antibody framework. The amino acid residues involved in antigen binding and derived from non-human antibody or antigen binding fragment thereof can be identified by checking the sequence and structure of animal monoclonal antibody variable region. The donor CDR framework amino acid residues which are different from the germ lines can be considered as being related. If it is not possible to determine the most closely related germ line, the sequence can be compared to the concensus sequence shared by subtypes or the animal antibody sequence with high similarity percentage. Rare framework residues are thought to be the result of a high mutation in somatic cells, and play an important role in binding.

[0043] In an embodiment of the present disclosure, the antibody or antigen-binding fragment thereof can further comprises a light chain constant region of human or murine κ, λ chain or variant thereof, or further comprises a heavy chain constant region of human or murine IgG1, IgG2, IgG3, IgG4 or variant thereof.

[0044] "Human antibody" and "antibody derived from human" can be used interchangeably, and can be antibodies derived from human or antibodies obtained from a genetically modified organism which has been "engineered" and produced by any method known in the art to produce specific human antibodies in response to antigen stimulation. In some technologies, elements of human heavy and light chain loci are introduced into cell lines of organisms derived from embryonic stem cell lines, and the endogenous heavy and light chain loci in these cell lines are targeted and disrupted. The targeted endogenous heavy and light chain loci included in these cell lines are disrupted. Transgenic organisms can synthesize human antibodies specific for human antigens, and the organisms can be used to produce hybridomas that secrete human antibodies. A human antibody can also be such antibody in which the heavy and light chains are encoded by nucleotide sequences derived from one or more human DNA sources. Fully human antibodies can also be constructed by gene or chromosome transfection methods and phage display technology, or constructed from B cells activated in vitro, all of which are known in the art.

[0045] "Monoclonal antibody" refers to an antibody obtained from a population of substantially homogeneous antibodies, that is, the individual antibodies constituting the population are identical and/or bind to the same epitope, except for possible variant antibodies (for example, variants containing naturally occurring mutations or mutations produced during the manufacture of monoclonal antibody preparations, and the mutations are usually present in minimal amounts). Unlike polyclonal antibody preparations that usually contain different antibodies directed against different determinants (epitopes), each monoclonal antibody of a monoclonal antibody preparation (formulation) is directed against a single determinant on the antigen. Therefore, the modifier "monoclonal" indicates the characteristics of the antibody obtained from a substantially homogeneous antibody population, and should not be interpreted as requiring any specific method to manufacture the antibody. For example, monoclonal antibodies used in accordance with the present disclosure can be prepared by various techniques, including but not limited to hybridoma methods, recombinant DNA methods, phage display methods, and methods by using transgenic animals containing all or part of human immunoglobulin loci. Such methods and other exemplary methods for preparing monoclonal antibodies are described herein.

[0046] The terms "full-length antibody", "full antibody", "whole antibody" and "complete antibody" are used interchangeably herein and refer to an antibody in a substantially complete form, as distinguished from antigen-binding fragments defined below. The term specifically refers to an antibody of which the heavy chain contains Fc region.

[0047] In addition, the VL domain and VH domain of the Fv fragment are encoded by two separate genes, however, they can be linked by a synthetic linker by using recombinant methods, to generate a single protein chain in which a monovalent molecular is formed by pairing the VL and VH domain (referred to as single chain Fv (scFv); see, e.g., Bird et al. (1988): 423-426; Science 242 and Huston et al (1988) Proc. Natl. Acad. Sci USA85:5879-5883). Such single chain antibodies are also intended to be included in the term of "antigen binding fragment" of an antibody. Such antibody fragments are obtained using conventional techniques known in the field, and are screened for functional fragments by using the same method as that for an intact antibody. Antigen binding portions can be produced by recombinant DNA technology or by enzymatic or chemical disruption of an intact immunoglobulin.

[0048] Antigen-binding fragments can also be incorporated into a single-chain molecule comprising a pair of tandem Fv fragments (VH-CH1-VH-CH1), and the pair of tandem Fv fragments forms a pair of antigen-binding regions together with complementary light chain polypeptides (Zapata et al., 1995 Protein Eng. 8(10): 1057-1062; and US Patent No. 5,641,870).

[0049] Fab is an antibody fragment obtained by treating an IgG antibody molecule with a papain (which cleaves the amino acid residue at position 224 of the H chain), and the antibody fragment has a molecular weight of about 50,000 Da and has antigen binding activity, in which about a half of the N-terminal side of H chain and the entire L chain are bound together through disulfide bond(s).

[0050] F(ab')2 is an antibody fragment having molecular weight of about 100,000 Da and having antigen binding activity and comprising two Fab regions which are bound at the hinge position, it can be produced by digesting the part downstream of the two disulfide bonds in the IgG hinge region with pepsin.

[0051] Fab' is an antibody fragment having a molecular weight of about 50,000 Da and having antigen binding activity, which is obtained by cleaving the disulfide bonds at the hinge region of the above-mentioned F(ab')2. Fab' can be produced by treating F(ab')2 that specifically recognizes and binds to an antigen with a reducing agent such as dithiothreitol.

[0052] Further, the Fab' can be produced by inserting DNA encoding Fab' of the antibody into a prokaryotic expression vector or eukaryotic expression vector and introducing the vector into a prokaryote or eukaryote to express the Fab'.

[0053] The term "single chain antibody", "single chain Fv" or "scFv" refers to a molecule comprising antibody heavy chain variable domain (or region; VH) connected to antibody light chain variable domain (or region; VL) by a linker. Such scFv molecules have general structure of NH2-VL-linker-VH-COOH or NH2-VH-linker-VL-COOH. Suitable linkers in the prior art consist of repeated GGGGS amino acid sequence or variant thereof, for example, variant with 1-4 (including 1, 2, 3 or 4) repeats (Holliger et al. (1993), Proc Natl Acad Sci USA. 90: 6444-6448). Other linkers usefull for the present disclosure are described by Alfthan et al. (1995), Protein Eng. 8:725-731, Choi et al. (2001), Eur J Immuno.31:94-106, Hu et al. (1996), Cancer Res.56:3055-3061, Kipriyanov et al. (1999), J Mol Biol. 293:41-56 and Roovers et al. (2001), Cancer Immunol Immunother. 50:51-59.

[0054] "Multispecific antibody" refers to an antibody that comprises two or more antigen-binding domains and is capable of binding two or more different epitopes (for example, two, three, four or more different epitopes), and the epitope can be present on the same or different antigens. Examples of multispecific antibodies include "bispecific antibodies" that bind to two different epitopes.

[0055] The term "bivalent bispecific antibody" of a tumor-associated antigen refers to a bispecific antibody, in which two antigen-binding regions are directed against a tumor-associated antigen target. For example, B7H3 bivalent bispecific antibody refers to the bispecific antibody comprising two antigen-binding regions targeting B7H3. The term "monovalent bispecific antibody" refers to a bispecific antibody, in which only one antigen-binding region is directed against a certain target. For example, B7H3 monovalent bispecific antibody refers to the bispecific antibody comprising one antigen-binding region targeting B7H3.

[0056] "Linker" or "linking fragment" refers to "L1" located between two protein domains for connecting the two domains, also refers to a connecting peptide sequence used to connect protein domains. It usually has a certain degree of flexibility, and the use of linkers will not cause the protein domain to lose its original functions.

[0057] Diabody is an antibody fragment wherein the scFv is dimerized, and it is an antibody fragment having bivalent antigen binding activity. In the bivalent antigen binding activity, the two antigens can be the same or different.

[0058] dsFv is obtained by substituting one amino acid residue in each of VH and VL with a cysteine residue, and then connecting the substituted polypeptides via a disulfide bond between the two cysteine residues. The amino acid residues to be substituted with a cysteine residue can be selected based on three-dimensional structure prediction of the antibody in accordance with known methods (Protein Engineering, 7, 697 (1994)).

[0059] In some embodiments of the present disclosure, the antigen-binding fragment can be produced by the following steps: obtaining cDNAs encoding the monoclonal antibody VH and/or VL of the present disclosure that specifically recognizes and binds to the antigen, and cDNAs encoding the other domains as required; constructing DNA encoding the antigen-binding fragment; inserting the DNA into a prokaryotic or eukaryotic expression vector, and then introducing the expression vector into a prokaryote or eukaryote to express the antigen-binding fragment.

[0060] "Fc region" can be a naturally occurring sequence or a variant Fc region. The boundaries of the Fc region of an immunoglobulin heavy chain are variable; however, the Fc region of a human IgG heavy chain is usually defined as a region extending from the amino acid residue at position Cys226 or from Pro230 to carboxyl terminus. The numbering of residues in the Fc region is according to the EU index numbering in Kabat. Kabat et al., Sequences of Proteins of Immunological Interest, 5th Edition Public Health Service, National Institutes of Health, Bethesda, Md., 1991. The Fc region of immunoglobulin usually has two constant domains, CH2 and CH3. Herein, the "first Fc" is also referred to as "Fc1", and the second Fc is also referred to as "Fc2".

[0061] In "V_a1-L1-V_b1-L2-V_c2-L2-V_d2-L4-Fc1" and "V_e3-L5-V_f3-L6-Fc2", V_a1, V_b1, V_c2, Vd2, Ve3 and V_f3 represent an antibody light chain variable region or heavy chain variable region, V_a1 and V_b1 bind to the first epitope of the antigen, V_c2 and Vd2 bind to the second epitope of the antigen, and Ve3 and V_f3 bind to the third epitope. The first epitope, the second epitope and the third epitope can be the same or not.

[0062] Similar to "VH_TAA-L1-VL_TAA-L2-VH_CD3-L3-VL_CD3-L4-Fc1", VH_TAA and VL_TAA represent an antibody variable region binding to an epitope of the tumor-associated antigen, and VH_CD3 and VL_CD3 represent an antibody variable region binding to an epitope of CD3.

[0063] In the present disclosure, "knob-Fc" refers to a knob-like spatial structure formed by incorporating a point mutation T366W in the Fc region of an antibody. Correspondingly, "hole-Fc" refers to a hole-like spatial structure formed by incorporating point mutations T366S, L368A, and Y407V in the Fc region of an antibody. Knob-Fc and hole-Fc are more likely to form heterodimers due to steric hindrance. In order to further promote the formation of heterodimers, point mutations S354C and Y349C can be introduced into knob-Fc and hole-Fc, respectively, to further promote the formation of heterodimers via disulfide bonds. Meanwhile, in order to eliminate or alleviate the ADCC effect caused by antibody Fc, substitution mutations of 234A and 235A can also be introduced into Fc. For example, the preferred knob-Fc and hole-Fc of the present disclosure are shown in SEQ ID NOs: 69 and 70, respectively. In a bispecific antibody, knob-Fc or hole-Fc can be used as either the Fc region of the first polypeptide chain or the Fc region of the second polypeptide chain. For a single bispecific antibody, Fc regions of the first and the second polypeptide chain can not both be knob-Fc or hole-Fc.

[0064] The term "amino acid difference" or "amino acid mutation" refers to the amino acid changes or mutations in a protein or polypeptide variant when compared to the original protein or polypeptide, and involves insertion, deletion or substitution of one or more amino acid(s) on the basis of the original protein or polypeptide.

[0065] "Variable region" of an antibody refers to an antibody light chain variable region (VL) or antibody heavy chain variable region (VH), alone or in combination. As known in the field, each of the heavy and light chain variable regions consists of three complementarity determining regions (CDRs) (also named as hypervariable regions) connected to four framework regions (FRs). The CDRs in each chain are held tightly together by FRs and contribute to the formation of an antigen binding site of the antibody together with the CDRs from the other chain. There are at least two techniques for determining CDR: (1) a method based on cross-species sequence variability (i.e., Kabat et al. Sequences of Proteins of Immunological Interest, (5th edition, 1991, National Institutes of Health, Bethesda MD)); and (2) a method based on the crystallographic study of antigen-antibody complexes (Al-Lazikani et al., J. Molec. Biol. 273:927-948 (1997)). As used herein, CDRs can refer to those determined by either of or the combination of the two methods.

[0066] The term "antibody framework" or "FR region" refers to a part of the variable domain, either VL or VH, which serves as a scaffold for the antigen binding loops (CDRs) of this variable domain. Essentially, it is a variable domain without CDRs.

[0067] The term "CDR" refers to one of the six hypervariable regions present in the antibody variable domain that mainly contribute to antigen binding. One of the most commonly used definitions of the 6 CDRs is provided by Kabat E.A. et al. ((1991) Sequences of proteins of immunological interest. NIH Publication 91-3242). As used in some embodiments herein, CDRs can be defined according to Kabat criteria (Kabat et al. Sequences of Proteins of Immunological Interest, (5th edition, 1991, National Institutes of Health, Bethesda MD)), for the definition of the light chain variable domains CDR1, CDR2 and CDR3 (LCDR1, LCDR2 and LCDR3), and the heavy chain variable domains CDR1, CDR2 and CDR3 (HCDR1, HCDR2 and HCDR3), for example, for the definition of CD3 antibody CDRs of the present disclosure. In other embodiments, CDRs can also be defined according to IMGT criteria and the like. For example, the B7H3 antibody CDRs are defined according to IMGT criteria.

[0068] The term "tumor antigen" refers to a substance produced by tumor cells, optionally a protein, including "tumor-associated antigen" or "TAA" (which refers to a protein that is produced in tumor cells and is differentially expressed in cancers versus the corresponding normal tissues) and "tumor specific antigen" or "TSA" (which refers to a tumor antigen that is produced in tumor cells and is specifically expressed or abnormally expressed in cancers compared to the corresponding normal tissues).

[0069] Non-limiting examples of "tumor-associated antigen" include, for example, AFP, ALK, B7H3, BAGE protein, BCMA, BIRC5(survivin), BIRC7, β-catenin, brc-ab1, BRCA1, BORIS, CA9, CA125, carbonic anhydrase IX, caspase -8, CALR, CCR5, CD19, CD20(MS4A1), CD22, CD30, CD33, CD38, CD40, CD123, CD133, CD138, CDK4, CEA, Claudin 18.2, cyclin-Bl, CYP1B1, EGFR, EGFRvIII, ErbB2/Her2, ErbB3, ErbB4, ETV6-AML, EpCAM, EphA2, Fra-1, FOLR1, GAGE protein (such as GAGE-1, -2), GD2, GD3, GloboH, Glypican-3, GM3, gp100, Her2, HLA/B-raf, HLA/k-ras, HLA/MAGE-A3, hTERT, IL13Rα2, LMP2, κ-Light, LeY, MAGE protein (such as MAGE-1, -2, -3, -4, -6 and -12), MART-1, mesothelin, ML-IAP, MOv-y, Muc1, Muc2, Muc3, Muc4, Muc5, Muc16(CA-125), MUM1, NA17, NKG2D, NY-BR1, NY-BR62, NY-BR85, NY-ESO1, OX40, p15, p53, PAP, PAX3, PAX5, PCTA-1, PLAC1, PRLR, PRAME, PSMA (FOLH1), RAGE protein, Ras, RGS5, Rho, ROR1, SART-1, SART-3, STEAP1, STEAP2, TAG-72, TGF-β, TMPRSS2, Thompson-nouvelle antigen (Tn), TRP-1, TRP-2, tyrosinase, uroplakin-3 and 5T4 (Trophoblast glycoprotein).

[0070] "CD3" refers to an antigen expressed on T cells as part of a multi-molecule T cell receptor (TCR), and it is a homodimer or heterodimer formed by two of the following four receptor chains: CD3-ε, CD3-δ, CD3-ζ and CD3-γ. Human CD3-ε (hCD3ε) comprises amino acid sequence described in UniProtKB/Swiss-Prot: P07766.2. Human CD3-δ (hCD3δ) comprises amino acid sequence described in UniProtKB/Swiss-Prot: P04234.1. Therefore, the term "CD3" refers to human CD3, unless specifically indicating that it is from a non-human species, such as "murine CD3", "monkey CD3", etc.

[0071] "Epitope" or "antigenic determinant" refers to a site on an antigen to which an immunoglobulin or antibody specifically binds. Epitopes usually include at least 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, or 15 consecutive or non-consecutive amino acids in a unique spatial conformation. See, for example, Epitope Mapping Protocols in Methods in Molecular Biology, Vol. 66, G.E. Morris, Ed. (1996).

[0072] The term "specifically bind to", "selectively bind to", "selective binding" or "specific binding" refers to the binding of an antibody to a predetermined epitope on an antigen. Typically, the antibody binds with an affinity (KD) of less than about 10^-8M, for example, less than about 10^-9 M, 10^-10 M or 10^-11 M or even less.

[0073] The term "affinity" refers to the strength of the interaction between an antibody and an antigen at a single epitope. Within each antigenic site, the variable region of the antibody "arm" interacts with the antigen at multiple amino acid sites via weak non-covalent forces; the greater the interaction, the stronger the affinity. As used herein, the term "high affinity" of an antibody or antigen-binding fragment thereof (e.g., Fab fragment) generally refers to an antibody or antigen-binding fragment with K_D of 1E^-9M or less (e.g., K_D of 1E^-10M or less, K_D of 1E^-11M or less, K_D of 1E^-12M or less, K_D of 1E^-13M or less, K_D of 1E^-14 M or less, etc.).

[0074] The term "KD" or "K_D" refers to a dissociation equilibrium constant for particular antibody-antigen interaction. Typically, the antibody binds to an antigen with a dissociation equilibrium constant (KD) of less than about 1E^-8M, for example, less than about 1E^-9M, 1E^-10M or 1E^-11M or even less, for example, as determined by Surface Plasma Resonance (SPR) technology in Biacore instrument. The smaller the KD value, the greater the affinity is.

[0075] The term "nucleic acid molecule" refers to DNA molecules and RNA molecules. The nucleic acid molecule can be single-stranded or double-stranded, but preferably is double-stranded DNA. A nucleic acid is "operably linked" when it is placed into a functional relationship with another nucleic acid sequence. For instance, a promoter or enhancer is operably linked to a coding sequence if it affects the transcription of the sequence.

[0076] The term "vector" means a construct capable of delivering one or more target genes or sequences, and preferably, expressing them in a host cell. Examples of vectors include, but are not limited to, viral vectors, naked DNA or RNA expression vectors, plasmids, cosmids or phage vectors, DNA or RNA expression vectors associated with cationic coagulants, DNA or RNA expression vectors encapsulated in liposomes, and certain eukaryotic cells such as producer cells.

[0077] Methods for producing and purifying antibodies and antigen-binding fragments are well known in the art, for example, Antibodies: A Laboratory Manual, Cold Spring Harbor, New York, chapters 5-8 and 15. For example, mice can be immunized with antigen or fragment thereof, and the resulting antibodies can then be renatured, purified, and sequenced for amino acid sequences by using conventional methods well known in the art. Antigen-binding fragments can also be prepared by conventional methods. The antibodies or antigen binding fragments of the present disclosure are engineered to incorporate one or more human framework regions onto the CDR regions derived from non-human antibody. Human FR germline sequences can be obtained from website http://imgt.cines.fr, or from The Immunoglobulin Facts Book, 2001, ISBN 012441351, by aligning against IMGT human antibody variable germline gene database by MOE software.

[0078] The term "host cell" refers to a cell into which an expression vector has been introduced. Host cells can include bacterial, microbial, plant or animal cells. Bacteria that are easily transformed include members of enterobacteriaceae, such as Escherichia coli or Salmonella strains; Bacillaceae such as Bacillus subtilis; Pneumococcus; Streptococcus and Haemophilus influenzae. Suitable microorganisms include Saccharomyces cerevisiae and Pichia pastoris. Suitable animal host cell lines include CHO (Chinese Hamster Ovary cell line), HEK293 cells (non-limiting examples such as HEK293E cells), and NS0 cells.

[0079] The engineered antibodies or antigen-binding fragments can be prepared and purified by conventional methods. For example, the cDNA sequences encoding the heavy and light chains can be cloned and recombined into a GS expression vector. The recombinant immunoglobulin expression vector can be stably transfected into CHO cells. As an alternative prior art, mammalian expression systems can lead to glycosylation of antibodies, especially in the highly conserved N-terminal sites of the Fc region. Stable clones were obtained by expressing an antibody specifically binding to an antigen. Positive clones can be expanded in serum-free culture medium in bioreactors for antibody production. Culture medium, into which an antibody has been secreted, can be purified by conventional techniques. For example, purification can be performed on Protein A or Protein G Sepharose FF column comprising adjusting buffer. The nonspecific binding components are washed out. The bound antibody is eluted by pH gradient and antibody fragments are detected by SDS-PAGE, and then pooled. The antibodies can be filtered and concentrated using common techniques. Soluble mixtures and multimers can be effectively removed by common techniques, such as size exclusion or ion exchange. The resulting product is needed to be frozen immediately, such as at -70°C, or lyophilized.

[0080] "Administration" or "treatment," as it applies to an animal, human, experimental subject, cell, tissue, organ, or biological fluid, refers to contacting an exogenous pharmaceutical, therapeutic, diagnostic agent, or composition with the animal, human, subject, cell, tissue, organ, or biological fluid. "Administration" and "treatment" can refer, e.g., to therapeutic, pharmacokinetic, diagnostic, research, and experimental methods. The treatment of a cell encompasses contacting a reagent with the cell, as well as contacting a reagent with a fluid, where the fluid is in contact with the cell. "Administration" or "treatment" also means in vitro or ex vivo treatments, e.g., of a cell, with a reagent, diagnostic, binding compound, or with another cell. "Treatment", as it applies to a human, veterinary, or research subject, refers to therapeutic treatment, prophylactic or preventative measures, research and diagnostic applications.

[0081] "Treat" means to administer a therapeutic agent, such as a composition containing any of the compounds of the present disclosure, internally or externally to a patient having one or more disease symptoms for which the agent has known therapeutic activity. Typically, the agent is administered in an amount effectively to alleviate one or more disease symptoms in the patient or population to be treated, by inducing the regression of or inhibiting the progression of such symptom(s) by any clinically measurable degree. The amount of a therapeutic agent that is effective to alleviate any particular disease symptom (also referred to as the "therapeutically effective amount") may vary according to various factors such as the disease state, age, and body weight of the patient, and the ability of the drug to elicit a desired response in the patient. Whether a disease symptom has been alleviated can be assessed by any clinical measurement typically used by physicians or other skilled healthcare providers to assess the severity or progression status of that symptom. While the embodiment of the present disclosure (e.g., a treatment method or article of manufacture) may not be effective in alleviating the target disease symptom(s) in every patient, it should alleviate the target disease symptom(s) in a statistically significant number of patients as determined by any statistical test known in the art such as Student's t-test, chi-square test, U-test according to Mann and Whitney, Kruskal-Wallis test (H-test), Jonckheere-Terpstra-test and Wilcoxon-test.

[0082] "Amino acid conservative modification" or "amino acid conservative substitution" means that the amino acids in a protein or polypeptide are substituted by other amino acids with similar characteristics (such as charge, side chain size, hydrophobicity/hydrophilicity, backbone conformation and rigidity, etc.), such that the changes can frequently be made without altering the biological activity or other required characteristics (such as affinity and/or specificity to an antigen) of the protein or polypeptide. Those skilled in the art recognize that, in general, single amino acid substitution in non-essential regions of a polypeptide does not substantially alter the biological activity (see, e.g., Watson et al. (1987) Molecular Biology of the Gene, The Benjamin/Cummings Pub. Co., p. 224 (4th Ed.)). In addition, substitutions with structurally or functionally similar amino acids are less likely to disrupt biological activity. Exemplary conservative substitutions are set forth in the table below, "Exemplary Amino Acid Conservative Substitutions".

[0083] Exemplary Amino Acid Conservative Substitutions

Original residue	Conservative substitution
Ala(A)	Gly; Ser
Arg(R)	Lys; His
Asn(N)	Gln; His; Asp
Asp(D)	Glu; Asn
Cys(C)	Ser; Ala; Val
Gln(Q)	Asn; Glu
Glu(E)	Asp; Gln
Gly(G)	Ala
His(H)	Asn; Gln
Ile(I)	Leu; Val
Leu(L)	Ile; Val
Lys(K)	Arg; His
Met(M)	Leu; Ile; Tyr
Phe(F)	Tyr; Met; Leu
Pro(P)	Ala
Ser(S)	Thr
Thr(T)	Ser
Trp(W)	Tyr; Phe
Tyr(Y)	Trp; Phe
Val(V)	Ile; Leu

[0084] "Effective amount" or "effective dose" refers to the amount of a medicament, compound, or pharmaceutical composition necessary to obtain any one or more beneficial or desired results. For prophylactic applications, beneficial or desired results include elimination or reduction of risk, reduction of severity, or delay of the onset of the disease, including the biochemical, histological, and behavioral manifestations of the condition, its complications, and intermediate pathological phenotypes during the development of the condition. For therapeutic applications, beneficial or desired results include clinical results, such as reduction of the incidence of various conditions associated with target antigen of the present disclosure or improvement of one or more symptoms of the condition, reduction of the dosage of other agents required to treat the condition, enhancement of the efficacy of another agent, and/or delay of the progression of the condition associated with the target antigen of the present disclosure in patients.

[0085] "Exogenous" refers to substances produced outside organisms, cells, or humans according to circumstances. "Endogenous" refers to substances produced in cells, organisms, or human bodies according to circumstances.

[0086] "Homology" and "identity" are interchangeable herein and refer to the sequence similarity between two polynucleotide sequences or between two polypeptide sequences. When a position in both of the two sequences to be compared is occupied by the same base or amino acid monomer subunit, e.g., if a position in each of two DNA molecules is occupied by adenine, then the molecules are homologous at that position. The percentage of homology between two sequences is a function of the number of matching or homologous positions shared by the two sequences divided by the number of positions to be compared and then multiplied by 100. For example, when two sequences are optimally aligned, if 6 out of 10 positions in the two sequences are matched or homologous, then the two sequences are 60% homologous; if 95 out of 100 positions in the two sequences are matched or homologous, then the two sequences are 95% homologous. Generally, when two sequences are aligned, comparison is performed to give the maximum homology percentage. For example, the comparison can be performed by BLAST algorithm, in which the parameters of the algorithm are selected to give the maximum match between each sequence over the entire length of each reference sequence.

[0087] The following references relate to the BLAST algorithm frequently used for sequence analysis: BLAST algorithm (BLAST ALGORITHMS): Altschul, SF et al., (1990) J. Mol. Biol. 215:403-410; Gish, W. et al., (1993) Nature Genet. 3:266-272; Madden, TL et al., (1996) Meth. Enzymol. 266:131-141; Altschul, SF et al., (1997) Nucleic Acids Res. 25:3389-3402; Zhang, J. et al. (1997) Genome Res. 7:649-656. Other conventional BLAST algorithms such as those available from NCBI BLAST are also well known to those skilled in the art.

[0088] "Isolated" refers to a purified state, in which the designated molecule is substantially free of other biological molecules, such as nucleic acids, proteins, lipids, carbohydrates, or other materials, such as cell debris and growth medium. In general, the term "isolated" is not intended to mean the complete absence of these materials or the absence of water, buffers or salts, unless they are present in an amount that significantly interferes with the experimental or therapeutic use of the compound as described herein.

[0089] "Optional" or "optionally" means that the event or circumstance that follows may but does not necessarily occur, and the description includes the instances in which the event or circumstance does or does not occur. For example, "optionally contains 1-3 antibody heavy chain variable regions" means the antibody heavy chain variable region with specific sequence can be, but need not be, present.

[0090] "Pharmaceutical composition" refers to a mixture containing one or more compounds according to the present disclosure or a physiologically/pharmaceutically acceptable salt or produg thereof and other chemical components, such as physiologically/pharmaceutically acceptable carriers and excipients. The pharmaceutical composition aims at promoting the administration to an organism, facilitating the absorption of the active ingredient and thereby exerting a biological effect.

[0091] The term "pharmaceutically acceptable carrier" refers to any inactive substance suitable for use in a formulation for the delivery of antibodies or antigen-binding fragments. The carrier can be an anti-adhesive agent, adhesive agent, coating agent, disintegrating agent, filler or diluent, preservative (such as antioxidant, antibacterial or antifungal agent), sweetener, absorption delaying agent, wetting agent, emulsifier, buffer, and the like. Examples of suitable pharmaceutically acceptable carriers include water, ethanol, polyols (such as glycerol, propylene glycol, polyethylene glycol, and the like), dextrose, vegetable oil (such as olive oil), saline, buffer, buffered saline, and isotonic agent,such as sugars, polyols, sorbitol and sodium chloride.

[0092] The term "cancer", "cancerous" or "malignant" refers to or describes a physiological condition in mammals generally characterized by unregulated cell growth. Examples of cancer include, but are not limited to, carcinoma, lymphoma, blastoma, sarcoma, and leukemia or lymphoid malignancies. More specific examples of the cancer include squamous cell carcinoma, myeloma, small cell lung cancer, non-small cell lung cancer (NSCLC), head and neck squamous cell carcinoma (HNSCC), glioma, Hodgkin's lymphoma , Non-Hodgkin's lymphoma, diffuse large B-cell lymphoma (DLBCL), follicular lymphoma, acute lymphoblastic leukemia (ALL), acute myeloid leukemia (AML), chronic lymphocytic leukemia (CLL), chronic myeloid leukemia (CML), primary mediastinal large B-cell lymphoma, mantle cell lymphoma (MCL), small lymphocytic lymphoma (SLL), T-cell/histocyte-rich large B-cell lymphoma, multiple myeloma, myeloid leukemia-protein 1 (Mcl-1), myelodysplastic syndrome (MDS), gastrointestinal (tract) cancer, kidney cancer, ovarian cancer, liver cancer, lymphoblastic leukemia, lymphocytic leukemia, colorectal cancer, endometrial cancer, kidney cancer, prostate cancer, thyroid cancer, melanoma, chondrosarcoma, neuroblastoma, pancreatic cancer, glioblastoma multiforme, gastric cancer, bone cancer, Ewing sarcoma, cervical cancer, brain cancer, gastric cancer, bladder cancer, hepatocellular tumor, breast cancer, colon cancer, hepatocellular cancer (HCC), clear cell renal cell carcinoma (RCC), head and neck cancer, hepatobiliary cancer, central nervous system cancer, esophagus cancer, malignant pleural mesothelioma, systemic light chain amyloidosis, lymphoplasmacytic lymphoma, myelodysplastic syndrome, myeloproliferative tumor, neuroendocrine tumor, Merkel cell cancel, testicular cancer, and skin cancer.

[0093] "Inflammatory disorder" refers to any disease, disorder, or syndrome in which an excessive or unregulated inflammatory response results in excessive inflammatory symptoms, host tissue damage, or loss of tissue function. "Inflammatory disease" also refers to a pathological state mediated by the chemotaxis pooling of leukocytes or neutrophils.

[0094] "Inflammation" refers to a protective local response caused by tissue damage or destruction, and it is useful to destroy, alleviate or eliminate (isolate) harmful substances and injured tissues. Inflammation is significantly related to the chemotaxis pooling of leukocytes or neutrophils. Inflammation can be caused by pathogenic organisms and viruses, as well as non-infectious causes such as trauma, reperfusion after myocardial infarction, stroke, immune response to foreign antigens, and autoimmune response.

[0095] "Autoimmune disease" refers to any group of diseases in which tissue damage is related to the responses to body's own components mediated by humoral or cell immunity. Non-limiting examples of autoimmune diseases include rheumatoid arthritis, psoriasis, Crohn's disease, ankylosing spondylitis, multiple sclerosis, type I diabetes, hepatitis, myocarditis, Sjogren syndrome, autoimmune hemolytic anemia due to transplant rejection, vesicular pemphigoid, Graves disease, Hashimoto thyroiditis, systemic lupus erythematosus (SLE), myasthenia gravis, pemphigus, pernicious anemia, etc.

[0096] In addition, another aspect of the present disclosure relates to methods for immunodetection or determination of target antigens, reagents for immunodetection or determination of target antigens, methods for immunodetection or determination of cells expressing target antigens, and the diagnostic agents for diagnosing diseases associated with target antigen-positive cells, comprising the monoclonal antibodies or antibody fragments of the present disclosure that specifically recognize and bind to the target antigen as an active ingredient.

[0097] In the present disclosure, the method for detecting or measuring the amount of the target antigen can be any known method. For example, it includes immunoassay or immunodetection method.

[0098] The immunoassay or immunodetection method is a method of detecting or measuring the amount of an antibody or antigen with a labeled antigen or antibody. Examples of immunoassay or immunodetection methods include radioactive substance-labeled immunoantibody method (RIA), enzyme immunoassay (EIA or ELISA), fluorescence immunoassay (FIA), luminescence immunoassay, western blotting, physicochemical method, and the like.

[0099] The above-mentioned diseases related to the target antigen-positive cells can be diagnosed by detecting or measuring the target antigen-expressing cells using the antibodies or antibody fragments of the present disclosure.

[0100] Cells expressing the polypeptide can be detected by the known immunodetection methods, preferably by immunoprecipitation, fluorescent cell staining, immunotissue staining, and the like. In addition, the method such as fluorescent antibody staining method with the FMAT8100HTS system (Applied Biosystem) can be used.

[0101] In the present disclosure, samples to be detected or measured for the target antigen are not particularly limited, as long as they are possible to contain cells expressing the target antigen, such as tissue cells, blood, plasma, serum, pancreatic juice, urine, stool, tissue fluid or culture medium.

[0102] Dependent on the required diagnostic method, the diagnostic agent containing the monoclonal antibody or antibody fragment thereof of the present disclosure can also contain reagents for performing an antigen-antibody reaction or reagents for detecting the reaction. The reagents for performing an antigen-antibody reaction include buffers, salts and the like. The reagents for detection include agents commonly used in immunoassay or immunodetection methods, for example, a labeled secondary antibody that recognizes the monoclonal antibody, antibody fragment or conjugate thereof, and a substrate corresponding to the label.

[0103] The details of one or more embodiments of the present disclosure are set forth in the above specification. The preferred methods and materials are described below, although any method and material similar or identical to those described herein can be used in the practice or testing of the present disclosure. Through the specification and claims, other features, purposes and advantages of the present disclosure will become apparent. In the specification and claims, the singular forms include plural aspects unless the context clearly dictates otherwise. Unless otherwise defined explicitly herein, all technical and scientific terms used herein have the meaning commonly understood by those skilled in the art to which this disclosure belongs. All patents and publications cited in the specification are incorporated by reference. The following examples are presented to more fully illustrate the preferred embodiments of the present disclosure. These examples should not be construed as limiting the scope of the present disclosure in any way, and the scope of the present disclosure is defined by the claims.

Examples

Preparation and screening of antibodies

[0104] Methods of making monoclonal antibodies are known in the art. One of methods that can be used is the method as described in Kohler, G. et al. (1975) "Continuous Cultures Of Fused Cells Secreting Antibody Of Predefined Specificity," Nature 256:495-497 or the modified form thereof. Typically, monoclonal antibodies are generated in non-human species, such as mice. Generally, mice or rats are used for immunization, but other animals such as rabbits and alpacas can also be used. Antibodies are prepared by immunizing mice with immunogenic amounts of cells, cell extracts, or protein preparations containing human CD3 or other target antigens (such as human B7H3). The immunogen can be, but is not limited to, primary cells, cultured cell lines, cancerous cells, nucleic acids or tissues.

[0105] In one embodiment, monoclonal antibodies binding to the target antigen are obtained by using host cells that over-express the target antigen as an immunogen. Such cells include, for example, but are not limited to, human T cells, cells over-expressing human B7H3.

[0106] In order to monitor the antibody response, a small amount of biological sample (e.g., blood) can be obtained from the animal and is tested for the titers of antibodies against the immunogen. The spleen and/or some large lymph nodes can be removed and dissociated into single cells. If desired, spleen cells can be selected by applying the cell suspension to an antigen-coated plate or well (after the non-specific adherent cells are removed). B cells expressing membrane-bound antigen-specific immunoglobulins will bind to the plate and will not be washed away by the remaining suspension. Subsequently, the resulting B cells or all dissociated spleen cells can be fused with myeloma cells (for example, X63-Ag8.653 and cells available from Salk Institute, Cell Distribution Center, San Diego, CA). Polyethylene glycol (PEG) can be used to fuse spleen or lymphocytes with myeloma cells to form hybridomas. The hybridomas are then cultured in a selective medium (for example, hypoxanthine, aminopterin, thymidine medium, otherwise referred to as "HAT medium"). Subsequently, the resulting hybridomas are seeded on the plate by limiting dilution, and the production of antibodies that specifically bind to the immunogen is analyzed by using, for example, FACS (Fluorescence Activated Cell Sorting) or Immunohistochemistry (IHC) screening. Subsequently, the selected monoclonal antibody-secreting hybridomas are cultured in vitro (e.g., in a tissue culture flask or hollow fiber reactor) or in vivo (e.g., as ascites in mice).

[0107] As another alternative to cell fusion technology, Epstein-Barr virus (EBV) immortalized B cells can be used to prepare the monoclonal antibodies of the present invention. If necessary, the hybridoma is proliferated and subcloned, and the anti-immunogen activity of the supernatant is analyzed by traditional analysis methods (for example, FACS, IHC, radioimmunoassay, enzyme immunoassay, fluorescent immunoassay, etc.).

[0108] In another alternative, the monoclonal antibodies against target antigens (such as CD3, B7H3) and any other equivalent antibodies can be sequenced and prepared recombinantly by any method known in the art (for example, humanization, preparation of fully human antibodies using transgenic mice, phage display technology, etc.). In one embodiment, the monoclonal antibodies against the target antigens (e.g., CD3, B7H3) are sequenced and then the polynucleotide sequences are cloned into a vector for expression or proliferation. The sequence encoding the antibody of interest can be maintained in a vector in the host cell and then the host cell can be proliferated and frozen for later use.

[0109] The polynucleotide sequences of the anti-CD3 monoclonal antibodies and any other equivalent antibody can be used for genetic manipulation to produce "humanized" antibodies to improve the affinity or other characteristics of the antibodies. The general principle of humanized antibodies includes retaining the basic sequence of the antigen-binding portion of the antibody, while the remaining non-human portion of the antibody is replaced with a human antibody sequence. Four steps are generally used for humanizing monoclonal antibodies. These steps are as follows: (1) determining the nucleotide sequences and the putative amino acid sequences of the light and heavy chain variable domains of an original antibody; (2) designing the humanized antibody, that is, determining which antibody framework region will be used in the process of humanization; (3) actual humanization methods/techniques and (4) transfection and expression of the humanized antibodies. See, for example, US Patent Nos. US4816567, US5807715, US5866692, and US6331415.

1. Preparation and screening of B7H3 antibodies

[0110] B cells were isolated by using human PBMCs, spleen, and lymph node tissues and RNAs were extracted to construct a library of natural single-stranded phage antibodies. The constructed natural single-chain phage antibody library was packaged to form phage particles, which were screened by panning using the liquid phase method. The phage was associated with the biotinylated B7H3 liquid phase, and was then separated by streptavidin magnetic beads. In order to obtain a positive sequence that binds to human B7H3, biotinylated human B7H3 was used for panning. Several monoclonal colonies were picked up and packaged into phage single-chain antibodies for phage ELISA test. The monoclonal phages were tested for their ability to bind to human B7H3 and murine B7H3, respectively, and B7H3 antibodies were obtained after screening.

[0111] The B7H3-related antigen used for detection are shown as follows:
Human B7H3 antigen for detection
Commercially available product (SinoBiological cat# 11188-H08H)
The sequence is as follows:

Note: The underlined portion represents the extracellular region of B7H3; the italics represent His-tag.
Monkey B7H3 antigen for detection
Commercially available product (SinoBiological cat#90806-C08H)
The sequence is as follows:



Note: The underlined portion represents the extracellular region of B7H3; the italics represent His-tag.
Mouse B7H3 antigen for detection
Commercially available product (SinoBiological cat# 50973-M08H)
The sequence is as follows:

Note: The underlined portion represents the extracellular region of B7H3; the italics represent His-tag.
Human B7H3 full-length amino acid sequence

Note: The double underlined portion represents the signal peptide (Signal peptide: 1-28); The underlined portion represents the extracellular region of B7H3 (Extracellular domain: 29-466), wherein 29-139 refers to Ig-like V-type 1 Domain, 145-238 refers to Ig-like C2-type 1 Domain; 243-357 refers to Ig-like V-type 2 Domain, and 363-456 refers to Ig-like C2-type 2 Domain; The dotted line represents the transmembrane region (Transmembrane domain:467-487); The italics represent the intracellular region (Cytoplasmic domain:488-534).
Monkey B7H3 full-length amino acid sequence

Note:
The double underlined portion represents the signal peptide (Signal peptide: 1-28);
The underlined portion represents the extracellular region of B7H3 (Extracellular domain: 29-466), wherein 29-139 refers to Ig-like V-type 1 Domain, 145-238 refers to Ig-like C2-type 1 Domain; 243-357 refers to Ig-like V-type 2 Domain, and 363-456 refers to Ig-like C2-type 2 Domain; The dotted line represents the transmembrane region (Transmembrane domain:467-487); The italics represent the intracellular region (Cytoplasmic domain:488-534).
>Mouse B7H3 full-length amino acid sequence

Note:
The double underlined portion represents the signal peptide (Signal peptide: 1-28); The underlined portion represents the extracellular region of B7H3 (Extracellular domain: 29-248); The dotted line represents the transmembrane region (Transmembrane domain:249-269); The italics represent the intracellular region (Cytoplasmic domain:270-316).
For B7H3 antibody h1702 obtained by screening, the sequences and CDR sequences defined by IMGT numbering criteria are as follows:

>h1702 VH

>h1702 VL

Note: Arranged in the order of FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4, the italic sequences represent FRs, and the underlined sequences represent CDRs.

Table 1 Light chain and heavy chain CDR sequences of B7H3 antibody h1702

Antibody	Heavy chain		Light chain
h1702	HCDR1	GFIFSSSA SEQ ID NO: 9	LCDR1	SGSVSTSHY SEQ ID NO: 12
	HCDR2	ISYDGSNK SEQ ID NO: 10	LCDR2	NTN SEQ ID NO: 13
	HCDR3	ARSARLYASFDY SEQ ID NO: 11	LCDR3	AIHVDRDIWV SEQ ID NO: 14

[0112] In order to further improve the performance of the bispecific antibodies, cysteine substitution mutations were carried out in the VH and VL of the B7H3 antibody h1702. Mutation G103C (according to natural amino acid sequence numbering, position 103 of SEQ ID NO: 16) was introduced into the light chain variable region, and mutation G44C (according to natural amino acid sequence numbering, position 44 of SEQ ID NO: 15) was introduced into the heavy chain variable region, such that a pair of disulfide bonds were formed. The heavy and light chain variable regions of the anti-B7H3 single-chain antibody after the mutation are as follows:

B7H3 VH44C:

B7H3 VL103C:

2. Preparation and screening of CD3 antibodies

[0113] Humanized CD3 antibodies can be obtained on the basis of murine CD3 antibodies by the methods such as mutation, library construction, humanization engineering and screening.

CD3 antigen related sequence information is as follows

[0114] 

Human CD3 antigen for detection

Commercially available product (SinoBiological cat# CT038-H2508H)

The sequences are as follows:

Human CD3ε (Human CD3ε)

Note:
The underlined portion represents the extracellular region of CD3ε (Extracellular domain: 23-126); the italics represent His tag.

Human CD3δ

Note:
The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-105); the italics represent the Flag tag.

Monkey CD3 antigen for detection

Commercially available product (Acro biosystem cat# CDD-C52W4-100ug)

The sequences are as follows:

Monkey CD3ε

Note:
The underlined portion represents the extracellular region of CD3ε (Extracellular domain: 22-117); the italics represent His tag.

Monkey CD3δ

Note:
The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-105); the italics represent the Flag tag.

Mouse CD3 antigen for detection

Commercially available product (SinoBiological cat# CT033-M2508H). The sequences are as follows:

Mouse CD3ε

Note:
The underlined portion represents the extracellular region of CD3ε (Extracellular domain: 22-108); the italics represent His tag.

Mouse CD3δ

Note:
The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-105); the italics represent the Flag tag.

Human CD3ε full-length amino acid sequence



Note:

The double underlined portion represents the signal peptide (Signal peptide: 1-28);

The underlined portion represents the extracellular region of CD3ε (Extracellular domain: 23-126), wherein 32-112 refers to Ig-like Domain; The dotted line represents the transmembrane region (Transmembrane domain: 127-152); The italics represent the intracellular region (Cytoplasmic domain: 153-207).

Human CD3δ full-length amino acid sequence

Note:

The double underlined portion represents the signal peptide (Signal peptide: 1-21);

The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-105); The dotted line represents the transmembrane region (Transmembrane domain: 106-126); The italics represent the intracellular region (Cytoplasmic domain: 127-171).

Monkey CD3ε full-length amino acid sequence

Note:

The double underlined portion represents the signal peptide (Signal peptide: 1-21);

The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-117); The dotted line represents the transmembrane region (Transmembrane domain: 118-138); The italics represent the intracellular region (Cytoplasmic domain: 139-198).

Monkey CD3δ full-length amino acid sequence

Note:

The double underlined portion represents the signal peptide (Signal peptide: 1-21);

The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-105); The dotted line represents the transmembrane region (Transmembrane domain: 106-126); The italics represent the intracellular region (Cytoplasmic domain: 127-171).

Mouse CD3ε full-length amino acid sequence

Note:

The double underlined portion represents the signal peptide (Signal peptide: 1-21);

The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-108); The dotted line represents the transmembrane region (Transmembrane domain: 109-134); The italics represent the intracellular region (Cytoplasmic domain: 135-189).

Mouse CD3δ full-length amino acid sequence

Note:

The double underlined portion represents the signal peptide (Signal peptide: 1-21);

The underlined portion represents the extracellular region of CD3δ (Extracellular domain: 22-105); The dotted line represents the transmembrane region (Transmembrane domain: 106-126); The italics represent the intracellular region (Cytoplasmic domain: 127-173).

[0115] After repeated analysis and optimization, a series of humanized anti-CD3 antibody sequences were obtained. The heavy chain variable region sequences are as follows:

Table 2 The heavy chain variable region sequence of CD3 humanized antibodies

Name	SEQ ID NO:	Sequence
HRH-1	29
HRH-2	30
HRH-3	31
HRH-4	32
HRH-5	33
HRH-6	34
HRH-7	35

[0116] The light chain variable region sequences are as follows: >HRL

Note: Arranged in the order of FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4, the italic sequences represent FRs, and the underlined sequences represent CDRs. The number and position of light and heavy chain variable region CDRs (LCDR1-LCDR3 and HCDR1-HCDR3) of the CD3 humanized antibodies listed herein and in table 3 below comply with the well-known Kabat numbering criteria.

Table 3 CDR sequences of CD3 antibodies

Antibody variable region	Heavy chain CDR		Antibody variable region	Light chain CDR
HRH-1	HCDR1	KYAMN SEQ ID NO: 37	HRL	LCDR1
	HCDR2			LCDR2	GTKFLAP SEQ ID NO: 49
	HCDR3	HGNFGNSYISYWAY SEQ ID NO: 39		LCDR3	VLWYSNRWV SEQ ID NO: 50
HRH-2	HCDR1	KYAMN SEQ ID NO: 37
	HCDR2
	HCDR3	HGNFGNPYISYWAY SEQ ID NO: 41
HRH-3	HCDR1	KYAMN SEQ ID NO: 37
	HCDR2
	HCDR3	HGNFGNEYISYWAY SEQ ID NO: 42
HRH-4	HCDR1	KYAMN SEQ ID NO: 37
	HCDR2
	HCDR3	HGNFGNSYISYWDY SEQ ID NO: 43
HRH-5	HCDR1	KYAMS SEQ ID NO: 44
	HCDR2
	HCDR3	HGNFGNSYISYWAY SEQ ID NO: 45
HRH-6	HCDR1	KYAMN SEQ ID NO: 37
	HCDR2
	HCDR3	HGNFGNSYISYWAY SEQ ID NO: 45
HRH-7	HCDR1	KYAMN SEQ ID NO: 37
	HCDR2
	HCDR3	HGNFGNSYISYWAY SEQ ID NO: 45

Construction and preparation of single chain antibodies

[0117] scFvs against B7H3 and scFvs against CD3 were generated by connecting the light and heavy chain variable regions derived from the above-mentioned B7H3 antibodies, and by connecting the light and heavy chain variable regions derived from the CD3 antibody respectively, wherein the linker can be selected from those well-known in the art. Exemplary linker can be selected from: (GGGGS)n or (GGGGS)n GGG, where n can be 1,2, 3, or 4.

[0118] Exemplary anti-B7H3 scFvs are as follows:

Table 4 Sequence listing of various anti-B7H3 single chain antibodies (scFvs)

Name of the Single chain antibody (structural form)	Sequence (SEQ ID NO:)
B7H3-scFv1 (VLB7H3-linker- VHB7H3)
B7H3-scFv2 (VHB7H3-linker- VLB7H3)
B7H3-scFv3(VLB7H3-linker-VHB7H3)
B7H3-scFv4(VHB 7H3-linker-VLB7H3)

[0119] Exemplary anti-CD3 scFvs are as follows:

Table 5 Sequence listing of various anti-CD3 single chain antibodies (scFvs)

Name of anti-CD3 scFvs (structural form)	Sequence (SEQ ID NO:)
CD3-scFv1H (VHCD3-linker-VLCD3)
CD3-scFv2H (VH CD3-linker-VL CD3)
CD3-scFv3H (VH CD3-linker-VL CD3)
CD3-scFv4H (VH CD3-linker-VL CD3)
CD3-scFv5H (VH CD3-linker-VL CD3)
CD3-scFv6H (VH CD3-linker-VL CD3)
CD3-scFv7H (VH CD3-linker-VL CD3)
CD3-scFv1L (VL CD3-linker-VHCD3)
CD3-scFv2L
(VL CD3-linker-VHCD3)
CD3-scFv3L (VL CD3-linker-VH CD3)
CD3-scFv4L (VL CD3-linker-VH CD3)
CD3-scFv5L (VL CD3-linker-VH CD3)
CD3-scFv6L (VL CD3-linker-VH CD3)
CD3-scFv7L (VL CD3-linker-VH CD3)

Construction and preparation of bispecific antibodies

B7H3 bivalent bispecific antibodies and B7H3 monovalent bispecific antibodies

[0120] In some embodiments of the present disclosure, the structure of the B7H3 bivalent bispecific antibodies is shown in Figure 1A, wherein the the C-terminus of the antibody can be or can not be labeled to His tag. Two B7H3 antigen-binding domains and one CD3 antigen-binding domain are configurated in a designed asymmetric structure of the two Fc-containing chains, wherein each B7H3 antigen-binding domain is on each of the two chains respectively, and the antigen-binding domains are all in the form of scFv. The Fc region can make the antibody maintain the normal half-life and favorable stability. The design of the two chains greatly reduces the probability of mismatches and improves the homogeneity of the sample and the yield of the target antibody. The specific molecular structure (Format) of the bispecific antibodies is shown in Table 6 below. In addition, the molecular structure of the B7H3 monovalent bispecific antibodies used in some embodiments of the present disclosure has Fc domain alone in the second polypeptide chain, without any antigen-binding domain, such structure is shown in Figure 1B.

Table 6 Structural representation of the bispecific antibodies

Name of the molecular structure	The arrangement order of the first polypeptide chain	The arrangement order of the second polypeptide chain
AFF1	VHB7H3-L1-VLB7H3-L2-VHCD3-L3-VLCD3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF2	VHB7H3-L1-VLB7H3-L2-VLCD3-L3-VHCD3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF3	VLB7H3-L1-VHB7H3-L2-VHCD3-L3-VLCD3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF4	VLB7H3-L1-VHB7H3-L2-VLCD3-L3-VHCD3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF5	VHCD3-L1-VLCD3-L2-VHB7H3-L3-VLB7H3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF6	VHCD3-L1-VLCD3-L2-VLB7H3-L3-VHB7H3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF7	VLCD3-L1-VHCD3-L2-VHB7H3-L3-VLB7H3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF8	VLCD3-L1-VHCD3-L2-VLB7H3-L3-VHB7H3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
AFF1-B	VHB7H3-L1-VLB7H3-L2-VHCD3-L3-VLCD3-L4-Fc1	VHB7H3-L5-VLB7H3-L6-FC2
AFF2-B	VHB7H3-L1-VLB7H3-L2-VLCD3-L3-VHCD3-L4-FC1	VHB7H3-L5-VLB7H3-L6-FC2
AFF3-B	VLB7H3-L1-VHB7H3-L2-VHCD3-L3-VLCD3-L4-FC1	VHB7H3-L5-VLB7H3-L6-FC2
AFF4-B	VLB7H3-L1-VHB7H3-L2-VLCD3-L3-VHCD3-L4-FC1	VHB7H3-L5-VLB7H3-L6-FC2
AFF5-B	VHCD3-L1-VLCD3-L2-VHB7H3-L3-VLB7H3-L4-FC1	VHB7H3-L5-VLB7H3-L6-FC2
AFF6-B	VHCD3-L1-VLCD3-L2-VLB7H3-L3-VHB7H3-L4-FC1	VHB7H3-L5-VLB7H3-L6-FC2
AFF7-B	VLCD3-L1-VHCD3-L2-VHB7H3-L3-VLB7H3-L4-FC1	VHB7H3-L5-VLB7H3-L6-FC2
AFF8-B	VLCD3-L1-VHCD3-L2-VLB7H3-L3-VHB7H3-L4-FC1	VHB7H3-L5-VLB7H3-L6-FC2
AF1	VHB7H3-L1-VLB7H3-L2-VHCD3-L3-VLCD3-L4-Fc1	Fc2
AF2	VHB7H3-L1-VLB7H3-L2-VLCD3-L3-VHCD3-L4-FC1	Fc2
AF3	VLB7H3-L1-VHB7H3-L2-VHCD3-L3-VLCD3-L4-FC1	Fc2
AF4	VLB7H3-L1-VHB7H3-L2-VLCD3-L3-VHCD3-L4-FC1	Fc2
AF5	VHCD3-L1-VLCD3-L2-VHB7H3-L3-VLB7H3-L4-FC1	Fc2
AF6	VHCD3-L1-VLCD3-L2-VLB7H3-L3-VHB7H3-L4-FC1	Fc2
AF7	VLCD3-L1-VHCD3-L2-VHB7H3-L3-VLB7H3-L4-FC1	Fc2
AF8	VLCD3-L1-VHCD3-L2-VLB7H3-L3-VHB7H3-L4-FC1	Fc2

Note: In this table, the carboxyl terminus of the first or the second polypeptide chain can be or can not be labeled to the His tag. L1, L2, L3, L4, L5, and L6 represent linkers for connecting each antigen-binding domain and the Fc region.

Table 7 Selection of linker sequence

Linker	Structure or sequence
L1	(GGGGS)n or (GGGGS)n GGG
L2	(GGGGS)n
L3	(GGGGS)n
L4	GGGDKTHTCPPCP (SEQ ID NO: 98)
L5	(GGGGS)n
L6	GGGDKTHTCPPCP (SEQ ID NO: 98)

[0121] Wherein n is selected from 1, 2, 3 or 4; preferably, n in L1 is 2 or 3, more preferably 3; n in L2 is 1 or 2, more preferably 1; n in L3 or L5 is 3. Optionally, the linker used to connect the antigen-binding domain and the Fc region can be selected from any other linker that can be used to connect the antibody functional domains, and is not limited to the linkers defined by the above sequences.

[0122] The Fc1 and Fc2 indicated in Table 6 above can be Fc with the same sequence, or can be knob-Fc and hole-Fc respectively, or hole-Fc and knob-Fc respectively. In some embodiments of the present disclosure, the sequences of knob-Fc and hole-Fc are preferably as shown in Table 8:

Table 8 Sequence listing of various Fcs

Name	SEQ ID NO	Sequence
knob-Fc	69
hole-Fc	70

[0123] For the above-mentioned light and heavy chain variable regions, single-chain antibodies, and bispecific antibodies, DNAs encoding the above-mentioned polypeptides or antigen-binding fragments can be constructed on the basis of the cDNAs encoding the VH and/or VL and other required domains, and the DNAs are inserted into prokaryotic expression vector(s) or eukaryotic expression vector(s), and then the expression vector(s) is(are) introduced into a prokaryotic or eukaryotic organism to express the polypeptides or antigen-binding fragments.

Example 1. Preparation of bispecific antibody molecules, positive control molecules, and negative control molecules

[0124] According to the method for designing bispecific antibody molecules of the present disclosure, specific bispecific antibody molecules were designed and prepared. Exemplary amino acid sequences of the molecules are shown in Table 9 below:

Table 9 Sequence listing of the bispecific antibodies

Antibody name (structure name)	The first polypeptide chain (SEQ ID NO)	The second polypeptide chain (SEQ ID NO)
113 (AFF3-2)
118 (AFF6-3)
119(AFF6-3L1D(GS))
126 (AFF6-2)
127 (AFF3-3)
128(AFF3-2(L2)2)
132 (AFF3-3SS)
131 (AFF3-1)
154 (AFF3-4)
156 (AFF3-6)
155 (AFF3-5)
177 (AFF3-7)
172 (AFF2-3)
171 (AFF1-3)
161 (AFF1-2)
162 (AFF2-2)
142 (AFF3-3B)
143 (AFF3-2B)
181 (AF3-1)	SEQ ID NO: 79	SEQ ID NO: 70
182 (AF3-2)	SEQ ID NO: 72
183 (AF3-3)	SEQ ID NO: 76
184 (AF3-4)	SEQ ID NO: 80
185 (AF3-5)	SEQ ID NO: 82
186 (AF3-6)	SEQ ID NO: 81
187 (AF3-7)	SEQ ID NO: 83

Note: The second polypeptide chains of the B7H3 bivalent bispecific antibody molecules 113, 118, 119, 126, 127, 128, 131, 132, 154, 155, 156, 161, 162, 171, 172 and 177 indicated in the above table are VL_B7H3-L5-VH_B7H3-L6-hole-Fc, as shown in SEQ ID NO: 71; and the second polypeptide chains of the B7H3 monovalent bispecific antibody molecules 181-187 are hole-Fc, as shown in SEQ ID NO: 70.

[0125] The amino acid sequences of the negative control (NC1, NC2, NC3) and the positive control (MGD009) bispecific antibodies used in this disclosure are as follows:

NC1: The B7H3 binding domain is replaced with a non-related antibody (anti-fluorescein antibody, anti-fluorescein), but the CD3 binding domain is retained. Reference literature for its amino acid sequence is: The anti-fluorescein antibody used to form the control DART diabody was antibody 4-4-20(Gruber. M. et al. (1994)).

Chain 1 (VH_CD3-VL_CD3-VL_ctrl-VH_ctrl-knob-Fc)

Chain 2 (VL _ctrl -VH _ctril-hole-Fc)

NC2: The B7H3 binding domain is retained, only the CD3 binding domain is replaced with a non-related antibody, anti-fluorescein.

Chain 1 (VH_ctrl-VL_ctrl-VL_B7H3-VH_B7H3-knob-Fc)

Chain 2



Note: The arrangement order is VL_B7H3-linker-VH_B7H3-linker-Fc. The underlined sequences represent the B7H3 antibody sequences, and the italics represent the hole-Fc sequence.

NC3

Chain 1 (VL_ctrl-VH_ctrl-VH_CD3-VL_CD3-knob-Fc-His tag)

Chain 2 (VL _ctrl -VH _ctrl -hole-Fc)

[0126] The positive control MGD009 comprises three chains, and the preparation and amino acid sequences can be found in the published patent application WO2017030926A1. Its amino acid sequences are as follows:

Chain 1 (B7H3VL-CD3VH-Fc)

Chain 2 (CD3VL-B7H3VH)

Chain 3 (Fc)

201 (DART-Fc Three-Chain Structure) 201 chain 1 (B7H3VL-CD3VH-E-Fc)

201 chain 2 (CD3VL-B7H3VH-K)

201 Chain 3

202 (Four-Chain Structure, where the mass ratio of the four chains Is Chain 1: Chain 2: Chain 3: Chain 4=1:2:1:1)

202 chain 1 (B7H3VH-CH1-Fc)

202 chain 2 (B7H3VL-CL)

202 chain 3 (B7H3 VH-CH1-CD3VH-CL)

202 chain 4 (CD3VL-CH1)

Example 2. Expression and purification of CD3-B7H3 bispecific antibodies

[0127] HEK293E cells were transfected with the plasmid expressing the bispecific antibody (chain 1: chain 2 is 1:1), the expression supernatant was collected 6 days later, and the impurities were removed by high-speed centrifugation. The clarified supernatant was purified on a Ni Sepharose excel column (GE Healthcare). The column was washed with PBS until the A280 reading dropped to the baseline, and then the column was washed with PBS+10mM imidazole to remove non-specifically bound impurity proteins, and the effluent was collected. Finally, the target protein was eluted with PBS solution containing 300mM imidazole, and the elution peaks were collected. The eluate samples were properly concentrated and then were further purified with the gel chromatography Superdex200 (GE) pre-equilibrated with 550 buffer (10mM acetic acid, pH5.5, 135mM NaCl). The target peak was collected. The sample was equilibrated against 559 buffer (10mM acetic acid, pH5.5, 9% sucrose) through a desalting column or ultrafiltration centrifuge tube, and aliquoted and stored at -80°C.

Test Example 1. Affinity of the bispecific antibodies to B7H3 and CD3 detected by BIAcore assay

[0128] The detection of antibody affinity to B7H3 and CD3 was performed in the form of capture antibody. BsAb was captured by CM5 biosensor chip (Cat. # BR-1005-30, GE) or Protein A (Cat. # 29127556, GE) biosensor chip coupled with Anti-Human IgG Antibody (Cat.#BR-1008-39, Lot. # 10260416, GE), and then each antigen flowed through the surface of the chip. The reaction signals were detected in real time with the Biacore T200 instrument to obtain the binding and dissociation curves. After the dissociation of each experimental cycle was completed, the chip was washed and regenerated with regeneration buffer Glycine1.5 (Cat# BR100354, GE) or 3 M MgCl₂ (from Human antibody capture kit, Cat. #BR100839, GE). The data were fitted with the (1:1) Langmuir model using GE Biacore T200 Evaluation version 3.0 software, and the affinity values were obtained.

[0129] With the arrangement order unchanged, the affinity of the bispecific antibodies to CD3 was slightly changed, when the sequences of CD3 antibody VH varried. When HRH-6 and HRH-5 sequences were used, the affinity of the antibody to CD3 was the weakest, and the binding to CD3 was not detectable by Biacore.

Table 10 Biacore assay results of the antigen-binding affinity of the bispecific antibodies with AFF3 structure

Bispecific antibody	CD3 VH involved	BIAcore KD(M)
131	HRH-1	4.07E-08
113	HRH-2	7.72E-08
127	HRH-3	9.72E-08
154	HRH-4	6.97E-08
156	HRH-6	No binding
155	HRH-5	No binding
177	HRH-7	1.62E-07

[0130] As an example, antibodies comprising HRH3 as the heavy chain variable region of the CD3 antigen-binding domain were selected for the assay. Among the selected antibodies, the test antibodies 118, 127 and 132 have an affinity to human B7H3 and human CD3 at levels of 10^-9 and 10^-8M, respectively, which are comparable to those of MGD009. These antibodies all have strong cross-binding activity to both monkey (cyno) B7H3 and human CD3 .

Table 11 Biacore assay results of the antigen-binding affinity of the bispecific antibodies comprising HRH3 arranged in different orders

Antibody	Affinity (M)
	Human B7H3	Monkey B7H3	Human CD3	Monkey CD3
MGD009	1.96E-09	2.65E-09	7.81E-08	4.94E-09
118	5.33E-09	7.31E-09	6.61E-08	1.59E-09
127	4.29E-09	5.61E-09	8.47E-08	1.24E-09
132	5.35E-09	6.90E-09	8.53E-08	8.13E-10

Test Example 2. Determination of the antibody binding ability at cell level

[0131] The ability of bispecific antibodies to bind to cell surface antigens was detected by FACS method. A498 (ATCC, HTB-44), CT26/hB7H3 (a recombinant cell line over-expressing human B7H3 in mouse cell CT26, constructed in-house, CT26 was obtained from the Cell Bank of Chinese Academy of Sciences, TCM37) and Jurkat recombinant cell lines (Jurkat cells were obtained from ATCC, PTS-TIB-152; the recombinant cell line was obtained on the basis of Jurkat cells by over-expressing luciferase gene and inserting NFAT response element upstream of the gene) were separately used for binding to antigens B7H3 and CD3 on cell surface.

[0132] FACS buffer (98% PBS, 2% FBS) was added into a 96-well U-shaped bottom plate (corning, 3795) to resuspend the cells, the serially diluted antibodies were added, incubated at 4°C for 1 hour, and the plate was washed with FACS buffer twice. Then APC anti-human IgG Fc Antibody (biolegend, Cat# 409306, dilution at 1:50) was added into each well, incubated at 4°C for 30 minutes, washed twice, the cells were resuspended in FACS buffer, and finally the fluorescence signal values were read by FACS CantoII (BD).

[0133] The results show that the B7H3 bivalent bispecific antibodies 118, 127, and 132 and the negative control antibody NC2 (in which the B7H3 binding domain was retained, and the CD3 binding domain was replaced with an unrelated antibody) are capable of binding to the A498 cell line (which highly expresses B7H3) (See Figure 2A), showing a gradient-dependent effect, with binding capacity stronger than that of MGD009, and specific for the B7H3 target. The negative control antibody NC1 (in which the B7H3 binding domain was replaced with a non-related antibody, but the CD3 binding domain was retained) does not bind to A498. Similarly, the bispecific antibodies 118, 127 and 132, MGD009 and NC2 strongly bind to CT26/hB7H3 (see Figure 2B), but do not bind to the CT26 cell line that does not express B7H3 (see Figure 2C), which also fully demonstrates that the test bispecific antibodies specifically bind to the B7H3 target on the cell membrane surface. The antibodies 118, 127 and 132 exhibit different binding ability from that of MGD009, such difference in binding ability is much significant on B7H3 over-expressing CT26/hB7H3 cell line than that on A498 cell line, indicating that the B7H3 bivalent bispecific antibodies have a more significant advantage for binding to B7H3 highly-expressing cells, and will have a better safety window compared to the B7H3 monovalent bispecific antibody MGD009.

[0134] The bispecific antibodies 118, 127 and 132 and the negative control antibody NC1 can bind to the Jurkat recombinant cell line (see Figure 2D), showing a gradient-dependent effect. Among these antibodies, 118 and NC1 have a Jurkat recombinant cell binding ability equivalent to that of MGD009, whereas 127 and 132 show a lightly weaker binding ability. This can be because that the CD3 binding domain is located between the B7H3 binding domain and FC, and a certain steric hindrance may affect the binding to Jurkat recombinant cells. The negative control antibody NC2 without the CD3 binding domain does not bind to Jurkat recombinant cells, indicating that the binding of the bispecific antibodies to Jurkat is specific for the CD3 target.

Test Example 3. In vitro PBMC killing assay

[0135] Bispecific antibody-mediated PBMC killing assay on tumor cells was achieved by quantitatively detecting cell proliferation. The content of ATP, an indicator of the metabolism of living cells and being directly proportional to the number of cells in the culture, was detected by using Cell Titer-glo in cells.

[0136] Four different target cells (T), including three tumor cell lines with different expression levels of B7H3 (A498, U87 (Cell Bank of Chinese Academy of Sciences, TCHu138), Detroit562 (ATCC, CCL-138)), and one negative control cell line CHOK1 that does not express B7H3 (ATCC, CCL-61) were used. Effector cells (E) were PBMCs obtained from healthy volunteers. The target cells were inoculated in a 96-well plate, cultured overnight, and equal amounts of freshly extracted PBMCs and serially diluted test bispecific antibodies (the highest final concentration was 300nM, diluted at 1:3), or PBS (control, with effector cells and target cells, without antibody) was added to each well on the next day. Blank controls (blank, medium only, without cells or antibodies) were set. The ratios of E:T were 10:1, 5:1, 5:1 and 5:1, respectively, for A498, U87, Detroit562 and CHOK1 cells. The cells were incubated for 48 hours and detected with Cell Titer-glo (refer to the instruction manual). The signal values were read on a microplate reader, and finally converted into the inhibition rate. The data were processed and analyzed by using Graphpad Prism 5.

3.1 Comparison of antibodies comprising different CD3 antigen-binding domain with varrying affinity

[0137] CD3 scFvs with different affinities were used to construct various bispecific antibodies, which show different in vitro target cell killing effects (see Figure 3A and Figure 3B). The bispecific antibodies 155, 156, 185 and 186 comprising HRH5 and HRH6, respectively show the weakest killing effect, which is consistent with the results of the Biacore affinity assay.

3.2 Comparison of B7H3 monovalent and bivalent bispecific antibodies

[0138] The comparison of the structure AFF3 (131 and 177 were used as exemplary antibodies of this structure) and AF3 (181 and 187 were used as exemplary antibodies of this structure) was performed as an illustrative example for the bispecific antibodies constructed from scFvs containing different anti-CD3 antibody heavy chain variable regions (see Fig. 4A and Fig. 4B). The B7H3 bivalent bispecific antibodies with AFF3 structure of CD3-B7H3 have significantly enhanced in vitro cell killing activity compared to the B7H3 monovalent bispecific antibodies with the AF3 structure. This applies for all bispecific antibodies containing different CD3 VHs.

Table 12 Antibody arrangement order

Name of the structure	First polypeptide chain	Second polypeptide chain
AFF3	VLB7H3-L1-VHB7H3-L2-VHCD3-L3-VLCD3-L4-Fc1	VLB7H3-L5-VHB7H3-L6-Fc2
AF3	VLB7H3-L1-VHB7H3-L2-VHCD3-L3-VLCD3-L4-Fc1	Fc2

3.3 The effect of different molecular structures of the B7H3 bivalent bispecific antibodies on tumor killing activity

[0139] The B7H3 bivalent bispecific antibody molecules 161, 162, 113 and 126 (see Figure 5A) and 113 and 143 (see Figure 5B) that share the same antigen-binding domain components but different arrangement orders, were tested in parallel for tumor cell killing activity. All the above molecules have HRH2 as the heavy chain variable region of the CD3 antigen-binding domain. The results show that the B7H3 bivalent specific antibody molecules with different arrangement orders all have significant killing effects on A498 cells. Among these antibodies, the 161, 162, 113 and 126 have a killing activity equivalent or slightly superior to that of MGD009. The arrangement order of the different structures has little effect on the tumor cell killing activity of the B7H3 bivalent bispecific antibodies.

Table 13 Comparison of the structure of the different test antibodies

Antibody	Name of the structure	First polypeptide chain	Second polypeptide chain
161	AFF1-2	VHB7H3-L1-VLB7H3-L2-VHCD3(HRH2)-L3-VLCD3-L4-Fc1	VLB7H3-L5-VHB7H3-L6-Fc2
162	AFF2-2	VHB7H3-L1-VLB7H3-L2-VLCD3-L3-VHCD3(HRH2)-L4-Fc1	VLB7H3-L5-VHB7H3-L6-Fc2
113	AFF3-2	VLB7H3-L1-VHB7H3-L2-VHCD3(HRH2)-L3-VLCD3-L4-Fc1	VLB7H3-L5-VHB7H3-L6-Fc2
126	AFF6-2	VHCD3(HRH2)-L1-VLCD3-L2-VLB7H3-L3-VHB7H3-L4-Fc1	VLB7H3-L5-VHB7H3-L6-Fc2
143	AFF3-2B	VLB7H3-L1-VHB7H3-L2-VHCD3(HRH2)-L3-VLCD3-L4-Fc1	VHB7H3-L5-VLB7H3-L6-Fc2

3.4 Bispecific antibodies have killing effects on tumor cell lines with different expression levels of B7H3

[0140] The three test bispecific antibodies 118, 127 and 132 were tested for the in vitro killing effects on A498, U87 and Detroit562 tumor cell lines. The killing effect is positively correlated with the expression level of B7H3. For example, the 118 has EC50 of 0.34, 2.4 and 14.5 nM for A498, U87 and Detroit562, respectively. All the three antibody molecules show this tendency. None of the bispecific antibodies have killing effect on the B7H3-negative control cell line CHOK1, and the negative control bispecific antibody NC1 did not have killing effect on any of the target cell lines. Together, these two aspects indicate that the cell killing is a target-specific killing, which requires redirecting the effector cells towards the B7H3 positive target cells by the bispecific antibodies.

Table 14. Redirection of PBMCs mediated by the test bispecific antibodies towards killing of different target cell lines

Cell line	B7H3 expres sion	118		127		132		NC1
		IC50 (nM)	Emax (%)	IC50 (nM)	Emax (%)	IC50 (nM)	Emax (%)	IC50 (nM)	Emax (%)
A498	high	0.34	65.9	2.4	67.5	2.94	66.4	>300	20.9
U87	interm ediate	2.4	67.2	8.2	69.5	11.9	67.6	>300	3.8
Detroit562	low	14.5	39.24	15.8	26.46	15.7	15.23	>300	0.95
CHOK1	Negati ve	>300	1.83	>300	3.91	>300	0.00	>300	0.81

3.5 Comparison of the killing effects of bispecific antibodies with different structures on A498 cells

[0141] The three test bispecific antibodies 127, 201 and 202 were tested for the in vitro killing effects on A498 tumor cell lines. The results show (see Figure 5C) that the bispecific antibodies with the three structures all have tumor killing activity, among which, the bispecific antibody 127 has a killing activity superior to that of 201 or 202.

Test Example 4. In vitro T cell activation assay

[0142] In order to detect the activation function of the bispecific antibodies on T cells, the expression of the NFAT-driven luciferase reporter gene after Jurkat's activation was measured using the Jurkat recombinant cell line, in the presence or absence of the A498 tumor cell line.

[0143] A498 cells were inoculated into a 96-well cell culture plate (1 × 10⁵/ml, 100 µL/well), and placed in a 37°C, 5% CO ₂ incubator for 20-24 h. On the next day, after the cell culture supernatant was removed, 90µl of Jurkat recombinant cell suspension (5.5×10 ⁵/ml) and 10µl of serially diluted test bispecific antibody (with the highest final concentration of 500nM, 1: 3 gradient dilution) were added into each well, and a negative control (which has A498 and Jurkat recombinant cells, without antibodies) and blank control (which has medium, without cells or antibodies) were set, and incubated at a 37°C, 5% CO₂ incubator for 5-6 hours. For the non-tumor cell-specific activation of Jurkat recombinant cells, Jurkat recombinant cells and the test antibodies were directly added into a blank 96-well culture plate. After the co-cultivation, 100µl of Bright-Glo Reagent (Bright-Glo™ Luciferase Assay System, Promega, Cat#: E2620) was added to each well, placed at room temperature for 5-10 minutes, and the chemiluminescence signal values were read on a multifunctional microplate. The fluorescence fold increase was calculated according to the formula:

4.1 B7H3 bivalent molecules with different arrangement orders can all effectively activate T cells

[0144] The B7H3 bivalent bispecific antibodies 118, 127, and 132 were tested for the activation of Jurkat recombinant cells in the presence or absence of A498 to verify the specific and non-specific activation effects of the bispecific antibodies on T cells. The results show that the B7H3 bivalent bispecific antibodies 118, 127 and 132 with different arrangement orders can effectively activate the Jurkat recombinant cell line and significantly induce the expression of luciferase, in the presence of the tumor cell line A498 (see Figure 6A). It demonstrates that the activation of Jurkat recombinant cells is specific to the B7H3 target, because the negative control antibody NC1 cannot induce the expression of luciferase. Co-recruitment of both Jurkat recombinant cells expressing CD3 and tumor cells expressing B7H3 through bispecific antibodies is required for the activation of Jurkat recombinant cells. In the case that Jurkat recombinant cells are present alone with the absence of A498 cells (see Figure 6B), the expression of luciferase is very low, and only a few weak signals can be detected at the several highest antibody concentration points.

Table 15 Antibody arrangement order

Antibody	First polypeptide chain	Second polypeptide chain
118	VHCD3(HRH3)-L1-VLCD3-L2-VLB7H3-L3-VHB7H3-L4-Fc	VLB7H3-L5-VHB7H3-L6-Fc
127	VLB7H3-L1-VHB7H3-L2-VHCD3(HRH3)-L3 -VLCD3-L4-Fc	VLB7H3-L5-VHB7H3-L6-Fc
132	VLB7H3-L1-VHB7H3-L2-VHCD3(HRH3)-L3-VLCD3-L4-Fc	VLB7H3-L5-VHB7H3-L5-Fc

4.2 Comparison of B7H3 monovalent and bivalent bispecific antibodies

[0145] The bivalent CD3-B7H3 bispecific antibodies have significantly enhanced target-specific T cell activation compared to the B7H3 monovalent bispecific antibodies, which is consistent with the enhanced in vitro tumor killing ability of the B7H3 bivalent molecules compared to the B7H3 monovalent molecules as indicated in Test Example 3. Meanwhile, the non-target-specific T cell activation remains unchanged. Therefore, the B7H3 bivalent molecule (131) has stronger efficacy than the B7H3 monovalent molecule (181) (see Figure 7A), whereas the side effects caused by the non-specific activation of T cells are not enhanced (see Figure 7B).

Table 16 Antibody structure

Antibody	First polypeptide chain	Second polypeptide chain
131	VLB7H3-L1-VHB7H3-L2-VHCD3-L3-VLCD3-L4-FC1	VLB7H3-L5-VHB7H3-L6-FC2
181	VLB7H3-L1-VHB7H3-L2-VHCD3-L3-VLCD3-L4-FC1	Fc2

Test Example 5. An In vitro cytokine secretion assay

[0146] The effector cells are redirected against the target cells under the mediation of the bispecific antibody, and release cytokines while killing the target cells. The cytokine secretion was analyzed by quantitatively detecting the content of the cytokines (including IL2, IFNγ, and TNFα) in the cell culture supernatant, by ELISA.

[0147] The experimental design and the antibodies used were the same as those described in Test Example 4. The cell culture supernatant was collected at the end of the in vitro killing assay and added into a 96-well plate (Corning#3795), and stored at -20°C for later use. For ELISA assay, the frozen culture supernatant was taken out, thawed at room temperature, centrifuged at 3500 rpm for 10 minutes, and the supernatant was collected for the ELISA assay. The procedures for ELISA followed the instructions supplied in the kit (Human IL-2 ELISA kit, Human IFN-γ ELISA Kit, Human TNF-α ELISA kit, Neobioscience, Cat # EHC003.96, EHC102g.96, EHC103a.96).

[0148] The results show that the test bispecific antibodies can effectively induce PBMC to secrete IL2, IFNγ and TNFα in the presence of both PBMC and B7H3-positive target cells A498 (see Figures 8A-8C), among these antibodies, MGD009 and 118 induce the highest secretion level of cytokines, followed by 127 and 132, and the negative control antibody NC1 induces the secretion of cytokine at a level beyond the range of detection sensitivity. MGD009 can significantly induce the release of IFNγ and TNFα at the three highest concentration points in the presence of both PBMC and B7H3-negative cell CHOK1 (see Figure 9A-Figure 9C), whereas the three test bispecific antibodies 118, 127 and 132 cannot induce the release of IFNγ and TNFα, indicating that the three test bispecific antibodies have better safety than MGD009, in terms of the secretion of non-target-specific cytokines.

Test Example 6. Pharmacodynamic test in mouse A498 model reconstructed with human PBMCs

[0149] In this test example, the anti-tumor efficacy of the test CD3-B7H3 bispecific antibodies of the present invention in mice was evaluated by using the NOG mouse (Beijing Charles River Experimental Animal Co., Ltd.) A498 model (ATCC) reconstructed with human PBMCs.

[0150] 5×10⁶ cells/mouse/100µl (containing 50% matrigel) of A498 cells were inoculated subcutaneously into the right flank of NOG mice. When the tumor volume in the tumor-bearing mice reached about 130-150mm³, the mice were randomly grouped, with 5-6 animals per group, and the day of grouping was defined as day 0 of the experiment. On day 0 or day 1, the PBMCs freshly extracted from two volunteers were mixed at a ratio of 1:1, and 5×10⁶ cells/100 µl was injected intraperitoneally in NOG mice, and each antibody was injected intraperitoneally, twice a week, a total of 6 doses. Tumor volumes and animal weights were monitored twice a week and data were recorded. Vehicle means a negative control group administrated with PBS buffer, instead of antibody.

[0151] Antibodies 118 and 119 showed certain anti-tumor efficacy at lower doses (Figure 10A), and exhibited a dose-dependent effect. Antibody 118 had tumor inhibition rate (TGI) of 22.17% and 60.39% at the doses of 0.01mpk and 0.03mpk, respectively, at the end of the experiment (day 20).

[0152] Antibody 113 showed a certain anti-tumor effect on day 14, and the tumor inhibition rates in 0.6mpk and 0.3mpk dose groups reached 70.05% (p<0.05) and 60.78% (p<0.05), respectively (Figure 10B). On 20 day, the anti-tumor effects kept increasing in a dose-dependent manner, and the anti-tumor rates are greater than 100% (p<0.001) and 77.92% (p<0.05), respectively.

[0153] Under the conditions of 0.12mpk and 0.36mpk dosed (Figure 10C), antibody 118 showed tumor inhibition rate of 39.18% and 57.44% (p<0.001) at the doses of 0.12mpk and 0.36mpk respectively on day 12, and the tumor inhibition rates reached 81.72% (p<0.01) and greater than 100% (p<0.001), respectively on day 21. Among the results, at the dose of 0.36mpk, one mouse even exhibited complete tumor regression (1/6).

[0154] At the dose of 0.36mpk (Figure 10D), antibody 126 had the tumor inhibition rate of 47.78% on day 21 (p<0.01). Antibody 128 showed a significant anti-tumor effect on day 19 (TGI=56.37%). By Day 21, the tumor-inhibition rate was increased to 69.28% (p<0.001). Antibody 127 exhibited a tumor inhibition rate of 76.20% on day 12 (p<0.001), and the tumor-inhibitory effect kept increasing on day 21, and the tumor inhibition rate is greater than 100% (p<0.001). In 3 out of the 5 animals, the tumor volumes were regressed compared to those on the day of grouping, and the tumor volumes in the other 2 animals were completely regressed.

[0155] The anti-tumor activity of antibody 127 was repeated in another experiment (Figure 10E). The tumor inhibition rate reached 90.6% on day 14 (p<0.001), and it is increased to 95.80% on day 17 (p< 0.001). 127 was still effective at lower dose (0.12mpk) and at lower administration frequency (once a week, 127-0.36mpk-qw), with the tumor inhibition rates on day 17 reached 51.37% (p<0.001) and 96.20% (p<0.001) respectively.

Test Example 7. Pharmacodynamic test in hCD3 KI mouse model

[0156] In this experiment, Balb/c-hCD3 mice were subcutaneously inoculated with CT26-hB7H3 tumor cell line (CT26 cells were derived from the Cell Bank of the Chinese Academy of Sciences, TCM37, and CT26-hB7H3 cells were obtained by expressing hB7H3) to evaluate the inhibitory effect of the CD3-B7H3 bispecific antibodies of the present invention on tumor growth in mice.

[0157] Female hCD3E Balb/c transgenic mice were purchased from Model Animal Research Center of Nanjing University (Certificate Number 201801374/5/6, license SCXK (Jiangsu) 2015-0001).

[0158] 8×10⁵ cells/mouse/100 µl of CT26-hB7H3 cells were inoculated subcutaneously into the right flank of hCD3 mice. When the tumor volumes in the tumor-bearing mice reached about 80-120 mm³, the mice were randomly divided into different groups, 7 mice per group. The day of grouping was defined as day 0 of the experiment, and the intraperitoneal injection of each antibody was performed, twice a week, for a total of 5 doses. The tumor volumes and animal weights were monitored twice a week and the data were recorded. Vehicle means a negative control group administrated with PBS buffer instead of the antibody.

[0159] The results show that antibody 118 showed strong efficacy after the initial administration at a dose of 1mpk (Figure 11A), and the tumor inhibition rate reached 38.34% on day 13 (p<0.05).

[0160] Antibody 132 had a tendency to inhibit tumor growth at a dose of 3.6 mpk (Figure 11B), and the tumor inhibition rate reached 26.35% on day 13.

Test Example 8. PK experiment in rats

[0161] In this experiment, the CD3-B7H3 bispecific antibodies were injected into the tail vein of SD rats, and the antibody concentrations in the rat's serum at different time points were detected to evaluate the metabolism of the CD3-B7H3 bispecific antibodies in SD rats.

[0162] The test drugs were injected into the tail vein of the rats with 3 mg/kg, and the administration volume was 5 mL/kg. Blood was collected at various time points, i.e., before administration and 5min, 8h, 1 d, 2 d, 4 d, 7 d, 10 d, 14 d, 21 d, 28 d after administration. The antibody concentrations in serum were detected by ELISA method. Two different ELISA methods were used, wherein B7H3 antigen (1µg/mL) or CD3 antigen (1µg/mL) were plated, and anti-human Fc-HRP (abcam, ab98624) was used as the secondary antibody. The pharmacokinetic parameters of the test drugs were calculated with Winnolin software, and the resulting main pharmacokinetic parameters are shown in Table 17.

[0163] The antibodies 118, 127 and 132 have a half-life of 4.9-8.1 days in terms of the B7H3 antigen-binding region, slightly longer than that of MGD009 and reaching the level of ordinary IgG antibodies; and a half-life of 3.2-5.6 days in terms of the CD3 antigen-binding region. Wherein, the kinetic parameters of two different antigen-binding regions for B7H3 and CD3 in antibody 118 are not much different, indicating that the integrity of the molecule in vivo is favorable, and the half-lives are 4.9 and 4.4 days, respectively. The antibody 127 has half-lives of 4.9 and 3.2 days, respectively, in terms of the two different antigen-binding regions for B7H3 and CD3. The differences in exposure amount and clearance rate are obvious, with the CD3 part inferior to the B7H3 part. This is more likely to be caused by weakening of the binding ability of CD3 rather than by molecular breakage, since the CD3 part is inside the molecular structure. Antibody 132 was obtained by incorporating a pair of disulfide bonds into B7H3 scFv, on the basis of the molecular sequence of antibody 127. This modification greatly increases the half-life of the molecule (65-75%), and also greatly improves the exposure amount and clearance rate.

Table 17 Main pharmacokinetic parameters in rats

Statistics	118		127		132		MGD009
	B7H3 Coated	CD3 Coated	B7H3 Coated	CD3 Coated	B7H3 Coated	CD3 Coated	B7H3 Coated
AUC 0-∞ (µ g/ml*h)	4797±603	3059±227	3328 ±171	437 ±37	5182±122	1634 ±654	2994±244
t1/2(h)	117±16	106±11	118 ±6	76 ±5	196 ±5	134 ±13	94.7±11.1
t 1/2 (d)	4.9±0.7	4.4±0.4	4.9 ±0.3	3.2 ±0.2	8.1±0.2	5.6 ±0.5	3.95±0.46

Claims

1. An antibody or antigen-binding fragment thereof specifically binding to human CD3, the antibody or antigen-binding fragment thereof comprising a heavy chain variable region and a light chain variable region, wherein:

the light chain variable region comprises LCDR1, LCDR2 and LCDR3 as shown in SEQ ID NOs: 48, 49 and 50, respectively, and

the heavy chain variable region is any one selected from the group consisting of the following i) to v):

i) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 38 and 39, respectively;

ii) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 40 and 41, respectively;

iii) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 40 and 42, respectively;

iv) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 40 and 43, respectively; and

v) a heavy chain variable region comprising HCDR1, HCDR2 and HCDR3 as shown in SEQ ID NOs: 37, 47 and 45, respectively.

2. The antibody or antigen-binding fragment thereof specifically binding to human CD3 according to claim 1, wherein the antibody or antigen-binding fragment thereof is a humanized antibody or antigen-binding fragment thereof.

3. The antibody or antigen-binding fragment thereof specifically binding to human CD3 according to claim 2, comprising a light chain variable region as shown in SEQ ID NO: 36, and/or
a heavy chain variable region as shown in any one selected from the group consisting of SEQ ID NOs: 29, 30, 31, 32, and 35.

4. The antibody or antigen-binding fragment thereof specifically binding to human CD3 according to claim 3, further comprising an antibody light chain constant region and/or a heavy chain constant region;
optionally, the light chain constant region is a light chain constant region of a human kappa, lambda chain or variant thereof, and the heavy chain constant region is a heavy chain constant region of a human IgG1, IgG2, IgG3, IgG4 or variant thereof.

5. The antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 4, wherein the antigen-binding fragment is selected from Fab, Fab', F(ab')2, dimerized V region (diabody) and disulfide-stabilized V region (dsFv).

6. A single-chain antibody, comprising the light chain variable region and the heavy chain variable region of the antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 3.

7. The single-chain antibody according to claim 6, wherein the sequence of the single-chain antibody is as shown in SEQ ID NO: 55, 56, 57, 58, 61, 62, 63, 64, 65 or 68.

8. A multispecific antibody specifically binding to human CD3 and tumor-associated antigen(s), the multispecific antibody comprising the single-chain antibody according to claim 6 or 7 or the antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 5.

9. The multispecific antibody according to claim 8, the tumor-associated antigen is selected from the group consisting of AFP, ALK, B7H3, BAGE protein, BCMA, BIRC5, BIRC7, β-catenin, brc-ab1, BRCA1, BORIS, CA9, CA125, carbonic anhydrase IX, caspase-8, CALR, CCR5, CD19, CD20, CD22, CD30, CD33, CD38, CD40, CD123, CD133, CD138, CDK4, CEA, Claudin 18.2, cyclin-Bl, CYP1B1, EGFR, EGFRvIII, ErbB2/Her2, ErbB3, ErbB4, ETV6-AML, EpCAM, EphA2, Fra-1, FOLR1, GAGE protein, GD2, GD3, GloboH, Glypican-3, GM3, gp100, Her2, HLA/B-raf, HLA/k-ras, HLA/MAGE-A3, hTERT, IL13Rα2, LMP2, κ-Light, LeY, MAGE protein, MART-1, Mesothelin, ML-IAP, MOv-γ Muc1, Muc2, Muc3, Muc4, Muc5, CA-125, MUM1, NA17, NKG2D, NY-BR1, NY-BR62, NY-BR85, NY-ESO1, OX40, p15, p53, PAP, PAX3, PAX5, PCTA-1, PLAC1, PRLR, PRAME, PSMA, RAGE protein, Ras, RGS5, Rho, ROR1, SART-1, SART-3, STEAP1, STEAP2, TAG-72, TGF-β, TMPRSS2, Thompson-nouvelle antigen, TRP-1, TRP-2, tyrosinase, uroplakin-3 and 5T4.

10. A pharmaceutical composition, comprising a therapeutically effective amount of the antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 5, the single chain antibody according to claim 6 or 7, or the multispecific antibody according to claim 8 or 9, and one or more pharmaceutically acceptable carriers, diluents, buffers or excipients.

11. An isolated nucleic acid molecule, encoding the antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 5, encoding the single-chain antibody of claim 6 or 7, or encoding the multispecific antibody of claim 8 or 9.

12. A recombinant vector, comprising the isolated nucleic acid molecule according to claim 11.

13. A host cell transformed with the recombinant vector of claim 12, said host cell is selected from prokaryotic cell and eukaryotic cell, preferably eukaryotic cell, more preferably mammalian cell or insect cell.

14. A method for producing the antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 5, the single-chain antibody according to claim 6 or 7, or the multispecific antibody according to claim 8 or 9, the method comprising culturing the host cell of claim 13 in a culture medium to form and accumulate the antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 5, the single-chain antibody according to claim 6 or 7, or the multispecific antibody according to claim 8 or 9, and recovering the antibody or antigen-binding fragment thereof, the single-chain antibody or the multispecific antibody from the culture.

15. The antibody or antigen-binding fragment thereof specifically binding to human CD3 according to any one of claims 1 to 5, the single-chain antibody according to claim 6 or 7, the multispecific antibody according to claim 8 or 9, the pharmaceutical composition according to claim 10, or the isolated nucleic acid molecule according to claim 11, as a medicament; preferably, the medicament is a medicament for the activation of T cells; more preferably, the medicament is a medicament for the treatment of cancer, autoimmune disease or inflammatory disease.

Drawing